Shear Transfer in Exposed Column Base Plates PDF

SHEAR TRANSFER IN EXPOSED COLUMN BASE PLATES
by
Ivan Gomez Chris Smith
Amit Kanvinde Gregory Deierlein

University of California, Davis Stanford University

Report Presented to the

American Institute of Steel Construction

March 2009
Executive Summary
This report presents the results of an experimental study investigating shear transfer
mechanisms of column base plate details. This investigation is the first phase of a broader
study whose aim is to develop improved design guidelines for column base plates. A
review of existing design guidelines and published research reveals that shear transfer
mechanisms in exposed base plates are not well understood, and methods to characterize
strength capacities in shear are not adequately supported by experimental data.
To investigate shear transfer in exposed column base plates, seven large scale specimens
were subjected to a combination of axial compression, axial tension and lateral shear
deformations. The tests investigated three mechanisms commonly used for shear transfer
in base plates, including (1) friction between the base plate and the grouted footing, with
and without steel shim stacks (2) anchor rod bearing and (3) shear key bearing. The base
plate tests are complemented by ancillary tests to characterize material properties.
Three tests which investigate surface friction indicate that the coefficient of friction
between a steel base plate with mill scale and a grouted surface, with or without shim
stacks, is 0.45. This value is lower than the design value of 0.55 recommended by the
Steel Design Guide One, published by the American Institute of Steel Construction.
Two tests that address anchor rod bearing indicate that without the consideration of rod
bending, the estimated strength of the anchor rod mechanism is significantly higher (i.e.
unconservative) as compared to the observed strength. The tests indicate that the anchor
rods initially bend in reverse curvature over a distance measured from the top of the grout
layer to the center of the plate washer welded to the top surface of the base plate.
However, this bending length increases as the grout sustains damage due to cyclic
inelastic loading. At larger lateral base plate slip, the anchor rods exhibit increased
strength due to second order geometric effects. In addition, the rods may come into
contact within the base plate holes, thus decreasing the bending length.
Two tests which investigate shear key bearing indicate that the current design method
used in the AISC Steel Design Guide One (based on the consideration of a uniform
tensile stress acting of a 45 degree failure cone) for calculating the side blowout strength
of a shear key loaded toward a free edge of an unreinforced concrete footing may be
unconservative for large foundation sizes (mean test-to-predicted ratios equal to 0.51 with
a coefficient of variation of 0.14). This unconservatism is attributed to the size effect in
concrete, which predicts that failure is controlled by the initiation of cracking, rather than
the development of concrete tensile strength over the failure surface. The concrete
capacity design (CCD) method, which incorporates this size effect from fracture
mechanics theory, is presented to characterize the concrete blowout strength. The
strength estimates determined as per this method agree closely with experimental
capacities (mean test-to-predicted ratio of 1.07 with a coefficient of variation of 0.19).
On-going work, as well as design strategies based on the experimental findings, is briefly
outlined.
ii
Acknowledgments
This project was funded by the American Institute of Steel Construction (AISC) and the
National Science Foundation (NSF; Grant Number NSF-CMMI 0421492). Trans Bay
Steel Corporation of Fairfield, California and PDM Steel Service Centers, Inc. of
Stockton, California generously donated steel materials and fabrication for this research
and their donations are gratefully acknowledged. The authors also thank Mr. Tom
Schlafly and Mr. Kurt Gustafson of AISC and Mr. Rick Drake of Fluor Corporation who
provided valuable advice and reviewed the testing plans. The large scale experiments
described in this report were conducted at the Network for Earthquake Engineering
Simulation (NEES) equipment site at the University of California at Berkeley in
Richmond, California. The authors are grateful to the site staff (Don Clyde, Shakhzod
Takhirov, Nathaniel Knight, David MacLam, Donald Patterson and Jose Robles) for
assistance during all phases of planning and testing. The authors would also like to thank
Jorge Camacho and Marshall Roberts from the University of California at Davis for
assisting with the test setup.
The findings and opinions of this report are those of the authors and do not necessarily
represent those of the major sponsors, i.e. AISC or NSF.
iii
Table of Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 General. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2. Background and Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.2 Current Design Provisions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3 Current State of Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.3.1 Surface Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3.2 Anchor Rod Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3.3 Shear Key Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4 Objectives of Current Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3. Experimental Program and Test Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2 Ancillary Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2.1 Anchor Rod Tension Tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2.2 Standard Concrete Test Cylinders. . . . . . . . . . . . . . . . . . . . . . . 16
3.2.3 Grout Test Cylinders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.3 Large Scale Experiments: Test Setup and Preparation . . . . . . . . . . . . . . 18
3.3.1 Test Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3.1.1 Test Setup for the Friction Tests:
Tests #1, #2 and #3. . . . . . . . . . . . . . . . . . . . . . . . . . 19
iv
3.3.1.2 Test Setup for the Anchor Rod Tests:
Tests #4 and #5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.3.1.3 Test Setup for the Shear Key Tests:
Tests #6 and #7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.3.2 Loading Protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.3.3 Instrumentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.4 Large Scale Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.4.1 Surface Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.4.2 Anchor Rod Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4. Summary and Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.1 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.1.1 Surface Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.1.2 Anchor Rod Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.2 Conclusion and Design Implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Appendix A – Ancillary Test Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Appendix B – Base Plate Test Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Appendix C – Anchor Rod Capacity Theory and Examples. . . . . . . . . . . . . . . . . . . . . 108
C.1 Development of Anchor Rod Capacity Equation Considering

Shear and Tension Loading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
v
C.2 Development of Anchor Rod Capacity Equation Considering
Shear and Tension Plus Flexural Loading. . . . . . . . . . . . . . . . . . . . . . . . 110
C.3 Variations on the Anchor Rod Capacity Equation. . . . . . . . . . . . . . . . . . 111
C.4 Anchor Rod Capacity Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Appendix D – Analysis of Anchor Rod Response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
D.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
D.2 Simplified Nonlinear Analysis of Anchor Rod Strength Under

Combined Axial Force, Bending and Shear. . . . . . . . . . . . . . . . . . . . . . . 116
D.3 Observations from Nonlinear Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Appendix E – Shear Key Capacity Theory and Examples . . . . . . . . . . . . . . . . . . . . . . 127
E.1 Development of the CCD Shear Capacity Equation. . . . . . . . . . . . . . . . 127
E.2 Example Calculation of Shear Key Capacity Using the CCD

Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
E.3 Example Calculation of Shear Key Capacity Using the 45

Degree Cone Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
vi
List of Tables
3.1 Summary of anchor rod tension tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2 Summary of concrete compression tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.3 Summary of grout compression tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.4 Base plate test matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.5 Base plate test loading details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.6 Base plate surface friction test results . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . 48
3.7 Anchor rod bearing test results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.8 Shear key bearing test results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
A.1 Anchor rod tension test results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
A.2 Concrete cylinder compressive strength test results. . . . . . . . . . . . . . . . . . . . . . . 94
A.3 Grout cylinder compressive strength test results. . . . . . . . . . . . . . . . . . . . . . . . . 95
C.1 Details of rod bending length. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
C.2 Details of anchor rod strength estimate calculations. . . . . . . . . . . . . . . . . . . . . . 114
E.1 Details of shear key capacity calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
vii
List of Figures
1.1 Schematic of typical exposed column base plate and associated shear
transfer mechanisms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.1 Concrete pedestal with anchor rods and shear key pocket . . . . . . . . . . . . . . . . . . 51
3.2 Base plate shear test setup (a) schematic (to scale; rear reaction wall
omitted for clarity) and (b) photograph. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.3 Steel shear loading assembly for base plate tests. . . . . . . . . . . . . . . . . . . . . . . . . 52
3.4 Steel shim stacks used for friction Tests #1 and #2. . . . . . . . . . . . . . . . . . . . . . . . 52
3.5 Shim stack positioning for (a) Test #1 and (b) Test #2. . . . . . . . . . . . . . . . . . . . . 53
3.6 Post-test photograph of grout pad for Test #3 . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.7 Schematic (a) plan view and (b) elevation view of reinforcing bar
details used in the concrete pedestals for the anchor rod tests #4 and
#5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.8 Schematic plan view of base plate specimen (to scale). . . . . . . . . . . . . . . . . . . . . 55
3.9 Welded plate washer detail (shown here for Test #5). . . . . . . . . . . . . . . . . . . . . . 55
3.10 Welded shear key (shown here for Test #6). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.11 Schematic detail of shear key geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.12 Schematic (a) plan view and (b) elevation view illustrating rebar details
in the concrete pedestals for the shear key tests (Test #6 and #7). . . . . . . . . . . . . 57
3.13 Typical lateral loading protocol for large scale surface friction and
anchor rod bearing tests (Tests #1 - #5). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
viii
3.14 Schematic illustrating instrument locations (to-scale). . . . . . . . . . . . . . . . . . . . . . 59
3.15 Representative response plot of the surface friction tests (shown here
for Test #3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.16 Condition of (a) the underside base plate, (b) the grout pad and (c) shim
stacks after Test #1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.17 Frictional force versus cumulative slip for Test #2. . . . . . . . . . . . . . . . . . . . . . . . 62
3.18 Response progression of each friction test illustrating the values

extracted for the coefficient of friction analysis. . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.19 Summary of friction tests illustrating the extracted data points used to
determine the coefficient of friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
3.20 Vertical uplift versus axial load during the initial axial tension loading
of anchor rod bearing Tests #4 and #5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
3.21 Anchor rod axial strains during initial axial uplift of base plate Test #4
(3/4” diameter rods). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.22 Vertical displacement of base plate versus cumulative lateral slip

during lateral loading of anchor rod bearing Tests #4 and #5 . . . . . . . . . . . . . . . 67
3.23 Load-displacement response of Test #4 – 3/4” diameter anchor rod. . . . . . . . . . 68
3.24 Load-displacement response of Test #5 – 1-1/4” diameter anchor rod. . . . . . . . 68
3.25 Post-test photographs showing (a) rod fracture of Test #4, (b) fractured
3/4” diameter rod from Test #4 and (c) fractured 1-1/4” diameter rod
from Test #5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.26 To scale illustration of fractured (a) 1-1/4” diameter anchor rod and (b)
3/4” rod relative to base plate, grout and concrete.. . . . . . . . . . . . . . . . . . . . . . . . 70
ix
3.27 Post-test photographs showing (a) grout damage of Test #4; (b) grout
damage and (c) concrete damage of Test #5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.28 Schematic illustrating various phenomenon in the anchor rod bearing

mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.29 Average axial strain measured from all anchor rods during the first
lateral displacement excursion to 0.8”. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.30 Schematic illustrating the mechanism causing the asymmetry in anchor

rod bearing load-deformation response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.31 (a) Superimposed photographs and (b) to-scale schematic indicating

as- built anchor rod positioning for Test #4 (3/4” diameter rods). . . . . . . . . . . . 75
3.32 (a) Superimposed photographs and (b) to-scale schematic indicating

as- built anchor rod positioning for Test #5 (1-1/4” diameter rods). . . . . . . . . . . 76
3.33 Anchor rod strength estimates and observed response for Test #4 (3/4”
diameter anchor rod); PV estimate out of range. . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.34 Anchor rod strength estimates and observed response for Test #5 (1-
1/4” diameter anchor rod); PV estimate out of range . . . . . . . . . . . . . . . . . . . . . . 77
3.35 Dimensions of the concrete pedestal with shear key. . . . . . . . . . . . . . . . . . . . . . 78
3.36 Representative load-displacement response for the shear key tests;

shown here for (a) the positive loading direction of Test #6 – 5.5”
shear key depth and (b) the negative loading direction. . . . . . . . . . . . . . . . . . . . . 79
3.37 Damage progression of initial loading of Test #6 at (a) 0.066” slip (b)
0.144” slip and (c) 0.246” slip (refer Figure 3.36a). . . . . . . . . . . . . . . . . . . . . . . 80
3.38 Post test photographs of Test #6 (a) elevation view (b) plan view. . . . . . . . . . . . 81
3.39 Failure surface of (a) Test #6 and (b) Test #7. . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
x
3.40 Schematic illustrating the effective stress area assumed by (a) the 45
degree cone method and (b) the CCD method. . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.41 Plot relating the shear capacity of a shear key embedded in concrete
( f c′ = 4000 psi) a distance “c” to a free edge using the 45 degree cone
method and the CCD method; experimental observations included. . . . . . . . . . . . 84
A.1 Stress-strain response of all anchor rod tension tests. . . . . . . . . . . . . . . . . . . . . . . 93
A.2 Photographs showing representative concrete cylinder test (a) before

and (b) after failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
A.3 Photographs showing representative grout cylinder test (a) before and
(b) after failure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
A.4 Grout product data from manufacturer (first page). . . . . . . . . . . . . . . . . . . . . . . . 96
A.5 Grout product data from manufacturer (second page). . . . . . . . . . . . . . . . . . . . . . 97
B.1 Schematic illustrating definitions of base plate test data. . . . . . . . . . . . . . . . . . . 98
B.2 Experimental response of Test #1 – Surface friction with shim stacks. . . . . . . . 99
B.3 Experimental response of Test #2 – Surface friction with shim stacks. . . . . . . . 100
B.4 Experimental response of Test #3 – Surface friction without shim

stacks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
B.5 Experimental response of Test #4 – 3/4” diameter anchor rod bearing . . . . . . . 102
B.6 Experimental response of Test #5 – 1-1/4” diameter anchor rod

bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
B.7 Experimental response of Test #6 – shear key bearing – 5.5”

embedment – positive loading direction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
xi
embedment – negative loading direction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

embedment – positive loading direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

embedment – negative loading direction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
B.11 Photograph showing post test condition (2nd load direction side
extracted) of pedestal from Test #6 – 5.5” shear key embedment depth
(a) elevation view of positive load direction side (b) elevation view of
negative load direction side (c) plan view . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
B.12 Photograph showing post test condition of pedestal from Test #7 –

3.0” shear key embedment depth (a) elevation view of positive load
direction side (b) elevation view of negative load direction side (c)
plan view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . 105
B.13 Photograph showing post test condition (failure surfaces extracted) of

pedestal from Test #7 – 3.0” shear key embedment depth (a) elevation
view of positive load direction side (b) elevation view of negative load
direction side (c) plan view. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
B.14 Contour map of Test #6 (5.5” shear key embedment depth - negative
load direction) failure surface recorded by HD laser scanner . . . . . . . . . . . . . . . 107
D.1 Schematic illustrating various phenomenon in the anchor rod bearing

mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
D.2 Free-body diagram of the deformed anchor rod. . . . . . . . . . . . . . . . . . . . . . . . . . . 124
D.3 Free-body diagram of half of the deformed anchor rod . . . . . . . . . . . . . . . . . . . . 125
xii
D.4 Experimental and predicted response of Test #4 (3/4” diameter anchor
rods) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
D.5 Experimental and predicted response of Test #5 (1-1/4” diameter

anchor rods) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
E.1 Schematic of concrete pedestal geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
E.2 Schematic illustrating the effective stress area for the CCD method
under various boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
E.3 Schematic illustrating the effective stress area for the 45 degree cone
method under various boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
xiii
List of Symbols
A - Nominal unthreaded area of the anchor rod
A35 - Effective area of the concrete failure surface based on a 35
degree projected plane from the bearing edges of the shear

lug to the free surface, excluding the bearing area of the
shear lug, inches2
A45 - Effective area of the concrete failure surface based on a 45
degree projected plane from the bearing edges of the shear

lug to the free surface, excluding the bearing area of the
shear lug, inches2
Ano - Projected area of the concrete failure surface for one
anchor at the concrete surface unlimited by edge influences

or neighboring anchors, idealizing the failure cone as a
pyramid with a base length of three times the embedment
depth
c - Free edge distance from the shear key, inches
COV - Coefficient of variation
d - Shear key embedment depth; or nominal (unthreaded)

diameter of anchor rod
du - Threaded diameter of anchor rod
xiv
E - Young’s modulus
e - Concrete footing corner distance
f c′ - Concrete compression strength, psi
ft , f v - Required tensile and shear stress, respectively
f ta - Required axial stress from axial loading
f tb - Required axial stress from bending
Fnt , Fnv - Nominal tensile strength and shear strength, respectively
Fu - Ultimate tensile strength of rod
Fu,rod - Mean ultimate strength of rod
Fy,rod - Mean yield strength of rod
h - Foundation height
hef - Effective concrete embedment depth, inches
It - Second moment of area of threaded anchor rod
k nc - Calibration factor; equal to 40 for cast-in situ headed studs
and headed anchor bolts
xv
k - Effective length factor of the anchor rod (i.e. lever arm
factor)
k 35 - Calibration constant; equal to 40/9 = 4. 4
l - Length of anchor rod in bending
Leffective - Effective bending length of anchor rod
Linitial - Effective bending length of the undeformed anchor rod
Ml - Resisting moment of anchor rod
n - Number of anchor rods
N no - Concrete cone failure load of a single anchor in uncracked
concrete unaffected by edge influences or overlapping

cones of neighboring anchors loaded in tension based on
the concrete capacity design (CCD) approach
P - Applied axial load
P1 - Resultant axial force on deformed anchor rod
R - Correlation coefficient
PV
Rmeasured ,eliptical - Predicted anchor rod capacity assuming rod tension and
shear; considering measured rod dimensions and an

elliptical stress interaction equation
xvi
PVM
Rmeasured ,eliptical - Predicted anchor rod capacity assuming rod tension, shear
and bending (bending length equal to the thickness of the

base plate plus half the thickness of the welded plate
washer); considering measured rod dimensions and an
PVM *
Rmeasured ,eliptical - Predicted anchor rod capacity assuming rod tension,
shear and bending (bending length equal to the thickness of

the base plate and grout pad plus half the thickness of the
welded plate washer) considering measured rod
dimensions and an elliptical stress interaction equation
PV
Rmeasured ,trilinear - Predicted anchor rod capacity assuming rod tension and
shear; considering measured rod dimensions and a

trilinear stress interaction equation
PVM
Rmeasured ,trilinear - Predicted anchor rod capacity assuming rod tension, shear

washer); considering measured rod dimensions and a
PVM *
Rmeasured ,trilinear - Predicted anchor rod capacity assuming rod tension,

dimensions and a trilinear stress interaction equation
xvii
RnoPVmin al ,eliptical - Predicted anchor rod capacity assuming rod tension and
shear; considering nominal rod dimensions and an

RnoPVM
min al ,eliptical - Predicted anchor rod capacity assuming rod tension, shear

washer); considering nominal rod dimensions and an
*
RnoPVM
min al ,eliptical - Predicted anchor rod capacity assuming rod tension,

dimensions and an elliptical stress interaction equation
RnoPVmin al ,trilinear - Predicted anchor rod capacity assuming rod tension and
shear; considering nominal rod dimensions and a

RnoPVM
min al ,trilinear - Predicted anchor rod capacity assuming rod tension, shear

washer); considering nominal rod dimensions and a
xviii
*
RnoPVM
min al ,trilinear - Predicted anchor rod capacity assuming rod tension,

dimensions and a trilinear stress interaction equation
R peak , R peak
reversed
- Peak lateral load observed for the anchor rod bearing tests
for each slip direction
tgrout - Thickness of the grout pad
tplate - Thickness of the base plate
twaher - Thickness of the welded plate washer
V - Applied shear load
V1 - Resultant shear force on deformed anchor rod
effective
V peak - Peak load observed during the second (i.e. final) load drop
minus the resistance offered by friction observed in the

shear key bearing tests
initial
V peak - Peak shear load prior to the first load drop observed in the
shear key bearing tests
ultimate
V peak - Peak shear load prior to the second load drop (i.e. ultimate
load) observed in the shear key bearing tests
w - Shear key width
xix
Z - Plastic section modulus of unthreaded anchor rod
Zt - Plastic section modulus of threaded anchor rod
Δ - Lateral displacement of anchor rod bolted end
θ - Effective angle of rod rotation
μ - Coefficient of friction
μmin - Minimum coefficient of friction value extracted from test

observation
σ failure - Nominal failure stress for CCD method
φ - Resistance factor
φ 45 - Strength reduction factor of the 45 degree cone method =
0.85
φCCD - Strength reduction factor of the CCD method = 0.75
xx
Chapter 1: Introduction
Chapter 1
Introduction
1.1 INTRODUCTION
This report describes the first phase of large scale testing conducted as part of a
comprehensive project whose aim is to characterize the performance of column base
plates under various loading conditions. The project proposal was initiated in response to
a request for proposal for proposal by the American Institute of Steel Construction (AISC
RFP 6807, 2006). At the time of writing this report, the project is ongoing, and future
reports will focus on other aspects of base plate response, including flexural loading. This
report focuses on completed work, which addresses shear-dominated loading.
1.2 GENERAL
Column base connections are critical components in steel structures because they transfer
axial forces, shear forces and moments to the foundation. Laboratory testing has
demonstrated the susceptibility of column base plate connections to various failure
modes, including weld fracture (e.g. Astaneh-Asl & Bergsma, 1993; Fahmy et al., 1999;
Burda & Itani, 1999), base plate yielding, anchor rod fracture and concrete crushing
(DeWolf & Sarisley, 1980). Recent studies by Grauvilardell et al. (2005) indicate that in
structural systems such as braced frames, a base plate connection may experience
extremely large shear-to-moment ratios, such that failure of the connection is dominated
by shear. However, experimental investigations of shear transfer in base plates are highly
limited, and most current design guidelines are based on adaptations of experimental data
from component testing in different contexts. For example, several studies investigate the
frictional behavior between steel and concrete/grout material interfaces. However, no
documented studies examine common detailing practice used for column erection, such
as the use of steel shim stacks for base plate construction. Similarly, most studies which
investigate anchor rods in base plates focus on concrete failure modes, rather than the
failure of anchor rods from axial tension, shear and bending. Thus, there is a lack of
1
experimental research which investigates structural details and modes of failure that may
be unique to the entire column base plate component. An investigation of these issues is
the primary motivation for the research program presented in this report.
Seven large scale experiments focusing on shear transfer in column base plates form the
main scientific basis of this study. These tests are supported by a set of analyses, as well
as ancillary tests of the materials used. The large scale tests were conducted at the state-
of-the-art Network for Earthquake Engineering Simulation (NEES) Structures Laboratory
at the University of California at Berkeley in Richmond, California. The experiments
were conducted using an elaborate testing apparatus which enabled the application of
combined tensile/compressive axial loading and cyclic lateral (shear) loading.
The large scale tests investigate three load transfer mechanisms which are commonly
used to resist shear in exposed column base plates, including surface friction, anchor rod
bearing and shear key bearing (see Figure 1.1). Three tests were conducted to investigate
surface friction, two tests investigated anchor rod bearing and two tests examined a shear
key detail.
Chapter 2 presents a detailed literature review and documents recent and past research on
column base plates, with a focus on shear transfer mechanisms. Chapter 2 evaluates this
research in the context of current design guidelines and common construction practice,
and presents the specific objectives of this study. Chapter 3 presents the experimental
data in detail, including the ancillary tests and the large scale base plate tests. The results
are analyzed and improved guidelines are presented to characterize the strength capacity
of various failure modes. Chapter 4 summarizes the findings of the study and presents
design implications, while outlining future work. Appendix A provides an inventory of
the ancillary tests, along with material descriptions. Detailed experimental data is
archived in Appendix B. The last three Appendices (C, D & E) outline proposed capacity
equations and provide calculation examples for the anchor rod bearing mechanism
(Appendix C and D) and the shear key bearing mechanism (Appendix E).
2
Steel
Column
Plate
Washer Base
& Nut Shear Force
Plate
Shim
Grout Stack
Concrete
Foundation
Anchor
Rods
Base Plate Detail Surface Friction
Shear Force Shear Force
Anchor Rod Bearing Shear Key Bearing
Figure 1.1 – Schematic of typical exposed column base plate and associated
shear transfer mechanisms (adapted from AISC Seismic Provisions, 2005)
3
Chapter 2: Background and Objectives
Chapter 2
Background and Objectives
2.1 INTRODUCTION
To provide background for the current investigation, this chapter presents a review of
literature relevant to shear transfer in exposed column base plates. Specifically, the
literature review addresses the three shear transfer mechanisms examined in this study,
i.e. surface friction, anchor rod bearing and shear key bearing. The chapter summarizes
currently used design strategies that incorporate these mechanisms, while outlining state-
of-the-art research relevant to improving these design guidelines. The chapter concludes
by describing the objectives of the current study in the context of the presented literature
review.
2.2 CURRENT DESIGN PROVISIONS

In 1990, the American Institute of Steel Construction (AISC) released a report authored
by DeWolf & Ricker (1990) which presented a compilation of existing information on
the design of base plates of steel columns. This report, part of the AISC’s Steel Design
Guide Series 1 – Column Base Plates, provides methods and suggestions for base plate
design, including bases subjected to axial compression and flexure. However, the report
cites the lack of design provisions for bases subjected to shear loading, in part due to
limited availability of test data. The report lists a limited number of publications
regarding shear transfer in base plates, such as Ballio & Mazzolani (1983), Goldman
(1983), Shipp & Haninger (1983) and Tronzo (1984), but recognizes that a large majority
of these are analytical studies.
The second edition of the AISC base plate design guide (Fisher & Kloiber, 2006) was
released partly in response to new significant research and a new Occupational Safety
and Health Administration (OSHA) provision requiring four anchor rods for most base
plate connections (OSHA, 2001). This new edition of the AISC Steel Design Guide 1
4
includes some additional information on shear transfer, such as providing a more

exhaustive treatment of anchor rod and shear key design for shear loading. However, not
all the additional information was supported by experimental verification of large scale
base plate prototypes. The Steel Design Guide 1 - Base Plate and Anchor Rod Design
(Fisher & Kloiber, 2006) relies on specifications from several organizations, including
the AISC Steel Construction Manual (2005), the American Concrete Institute (ACI)
Committee 318 (2002), and OSHA (2001). The Steel Design Guide 1, along with its
referenced codes and standards, form the basis of typical steel column base plate design
considerations in the United States.
Additionally, special design provisions have been developed for column base plates of
structures subjected to special loadings. For example, both the AISC Seismic Provisions
(AISC Seismic Design Manual, 2005) and the ACI code requirements for concrete
structures for nuclear facilities (ACI 349, 2006) provide guidelines for the design of
column base plates. In the context of shear transfer, the ACI 349 specifications provide
the most exhaustive and detailed design requirements. On the other hand, the
Commentary to the AISC Seismic Provisions addresses shear transfer in a more
qualitative sense while emphasizing that future experimental research on column base
plates should investigate seismic loading and larger prototypes subjected to shear loading.
2.3 CURRENT STATE OF RESEARCH

A recently published AISC sponsored synthesis of research on column base connections
(Grauvilardell et al., 2005) reveals that shear transfer in exposed column base plates are
not as well understood as flexural or axial load transfer. In fact, only two references listed
in this exhaustive synthesis, Goldman (1983) and Tronzo (1984), specifically address
shear transfer. These studies focus on design examples and design methods without a
detailed description of supporting experimental data. The authors of these two
publications express their concern for the lack of building code provisions and the
general lack of experimental research on base plate shear transfer, especially in the
context of column bases attached to braced frames, where the shear-to-moment ratio is
typically large.
5
An independent review of literature by the authors of the current study reveals that
experimental research of shear transfer in base plates has not been explicitly addressed,
although the associated mechanisms (e.g. surface friction or certain modes of concrete or
anchor rod failure) have been researched in separate contexts. Thus, current design
provisions for base plates typically adapt and combine findings from several of these
studies, which rely on small-scale component tests of specific failure mechanisms (e.g.
pullout or shear tests of individual anchor rods). Therefore, there is a lack of data
involving large scale base plate components, where various mechanisms may interact
with each other or may be influenced by the construction procedures or geometry of the
base connection itself. Moreover, few studies examine the effect of cyclic loading on
base plate shear transfer details, which is important from the perspective of seismic
design.
This section summarizes research relevant to the development of design considerations

for shear transfer in exposed column base plates. Specifically, this section outlines studies
that address three common shear transfer mechanisms, which are featured in the AISC
Steel Design Guide 1 (Fisher & Kloiber, 2006) and subsequently examined
experimentally in this study. These mechanisms include (1) surface friction (2) anchor
rod bearing and (3) shear lug (shear key) bearing. Other details have been proposed for
shear transfer in base plates. For example, the column and base plate can be embedded
directly into the concrete foundation or attached to a grade beam (Grauvilardell et al.,
2005).
For a more extensive literature review regarding column base plates, the authors refer the
reader to the first edition of the AISC Steel Design Guide 1 (DeWolf & Ricker, 1990) and
the aforementioned research synthesis (Grauvilardell et al., 2005).
2.3.1 Surface Friction

Under situations of low column base shear (e.g. in low-rise buildings subjected to modest
wind forces), shear loading in column base plates may be adequately resisted by surface
friction between the steel plate and grout/concrete interface due to the axial compressive
6
gravity load in the column. However, large lateral loads sustained by the structure
(especially in braced frames) may result in uplift of the base connection such that the
frictional resistance is reduced (or eliminated) and the shear force must be transferred to
the foundation by other means (e.g. through anchor rods or a shear key). In addition,
under seismic loading, high shear loads and other issues such as foundation rocking,
vibrations and tension uplift render surface friction an impractical choice for shear
transfer. Astaneh-Asl (2008) points outs that most seismic codes do not allow the surface
friction mechanism for base plates. Nevertheless, surface friction will always be active
when a compressive axial load is present and may be adequate to transfer low shear
forces.
The AISC Steel Design Guide 1 (unless otherwise noted, this reference henceforth
indicates the second edition of the AISC Steel Design Guide 1; i.e. Fisher & Kloiber,
2006) provides design coefficient of friction values for steel on grout and steel on
concrete (listed as 0.55 and 0.70, respectively). These values are adapted from ACI 349
criteria (ACI, 1985) and are based on historical test programs which have investigated
steel on grout/concrete surface friction under monotonic loading. Examples of such
studies investigating steel on grout friction include Cannon et al. (1975), Rabbat &
Russell (1985) and more recently Nagae et al. (2006). These studies report average
coefficient of friction values in the range of 0.52 - 0.65. Of these studies, only the most
recent (Nagae et al., 2006) examines frictional behavior between grout and steel under
reversed cyclic loading. This study provides a theoretical analysis of the surface friction
mechanism and, from small scale experiments, concludes that a friction coefficient of
0.50 is appropriate for design.
Although base plates typically rest on a grouted surface, surface friction between
concrete and steel has been studied by various researchers including Burdette et al.
(1987), Cook & Klingner (1991) and Baltay & Gjelsvik (1990). These studies examine
frictional behavior under monotonic loading and report average friction coefficient values
in the range 0.43 - 0.65. However, grout, rather than concrete, is typically installed at the
base plate interface, especially for large bases and those with shear keys.
7
It is important to note that none of the surface friction studies mentioned in this section
specifically investigate the interaction of column erection procedures, such as the use of
shim stacks, on the frictional response of base plates. In addition, while past tests
investigate a large range of surface pressures (e.g. 1-10,000 psi for Baltay & Gjelsvik
[1990]; 460-1,400 psi for Nagae et al. [2006]), most experimental programs investigate
relatively small scale base plates. For example, the base plates used in the study of Nagae
et al. (2006) are only 10 inches square. Nevertheless, the study publication by Nagae et
al. (2006) provides a detailed theoretical analysis of surface friction between steel and
grout. A combined theory of adhesion and plowing, originally used for friction between
steels, is applied to interpret the friction behavior and indicates reasonable agreement
with test results.
In summary, coefficient of friction values for the design of column base plate shear
transfer are, in general, based on studies that may not incorporate all the effects observed
in actual column base plate details. In addition to the limited large scale experimental
data of friction between steel and grout (especially under cyclic loading), other issues not
addressed by these studies include the effect of column erection procedures (e.g. the use
of shim stacks) and the associated evolution of frictional response.
2.3.2 Anchor Rod Bearing

Shear forces in column bases may be resisted by the base plate bearing against anchor
rods embedded in the concrete foundation. This mechanism may be practical under low
compressive axial loads or under axial tension (i.e. uplift) at the column base, when
sufficient frictional resistance cannot be developed. Fisher & Kloiber (2006) note that the
use of anchor rods to transfer shear forces must be carefully examined due to several
assumptions regarding the force transfer to the rods. A primary concern is the uncertainty
of uniform transfer of shear loads to all the anchor rods due to lateral displacement (or
slip) of the base plate. Highly oversized anchor rod holes (larger than standard oversize
holes – refer Fisher & Kloiber, 2006) are typically used in base plates in order to
compensate for construction tolerances. To reduce the extent of slip before engagement
8
of all anchor rods, plate washers (installed around the rods, with a minimal clearance)
may be fillet welded to the top surface of base plate.
As discussed above, anchor rods may be used to resist axial tension and shear loads, or a
combination thereof. Consequently, the AISC Steel Design Guide 1 provides guidelines
and design examples to assist in the design of anchor rods to resist shear and tension.
Transfer of shear through anchor rod bearing involves several issues. Summarized
briefly, these include (1) failure of either the grout or the concrete in the immediate
vicinity of the anchor rod either due to pullout, localized crushing or free edge blowout
(under shear), (2) the effects associated with eccentric placement of anchor rods within
the holes of the base plate, which may result in nonuniform load distribution amongst the
anchor rods and (3) failure of the anchor rod itself due to a combination of axial and
shear loading, which may also produce flexure (i.e. bending) in the rods. An additional
consideration, not addressed by the Steel Design Guide 1, is the response related to
second order effects due to large lateral base plate displacements, which may increase the
shear strength and stiffness of the connection due to increased tension forces in the
anchor rods.
With respect to issue (1) above, various researchers have examined the strength of steel
embedments and anchors in concrete, mainly in the context of the design of concrete
systems. These include the studies of Conrad (1969), Cannon et al. (1975), Bailey &
Burdette (1977) and Cannon et al. (1981), which all consider anchor rods embedded in
concrete loaded in monotonic shear or tension. Klinger et al. (1982) investigated anchor
bolts subject to monotonic and reversed cyclic shear loads and provided design equations
and design recommendations. However, it is important to note that these studies primarily
investigate concrete failure modes. Thus, the design considerations from these studies
focus on issues such as providing adequate anchor embedment lengths, edge distances
and reinforcement details. Several of these considerations have been adopted into
concrete design codes (e.g. ACI 318 and ACI 349). Building on these experimental
programs, Fuchs et al. (1995) developed the concrete capacity design (CCD) method
9
which is currently the preferred approach for the concrete design of embedded fasteners
under shear or tensile loading (ACI 349, 2006).
With respect to issue (2) above, the installation of welded plate washers typically ensures
that all anchor rods are engaged by the base plate almost simultaneously. However, the
eccentric placement of anchor rods may have other consequences. For example, anchor
rods that are close to the edge of the base plate hole in the direction of loading may
exhibit restrained bending (due to interaction with the base plate), thereby influencing
their strength and mode of failure. This issue is discussed later in Chapter 3.
With respect to issue (3) above, the AISC Steel Design Guide 1 suggests that anchor rods
in base plates be designed with adequate consideration of the interaction of axial tension,
flexural and shear stresses. The Design Guide refers to an equation listed in the AISC
Steel Construction Manual (section J3.7 in the commentary; AISC, 2005) for a
description of the interaction relationship between axial (normal) stresses and shear
stresses. This interaction relationship is based on extensive studies by Kulak et al. (1987).
The Design Guide suggests that for the calculation of flexural stresses in the anchor rods,
it is assumed that the anchor rods bend freely in double curvature over a certain length,
which is measured from the top of the grout pad to the center of welded plate washer (i.e.
the base plate thickness plus half the plate washer thickness).
Response of anchor rods in grouted base plates subjected to combined monotonic lateral
and tensile loading was studied experimentally by Adihardjo and Soltis (1979). This
study determined that existing shear/tension interaction equations based on bolts
embedded directly into concrete (from PCI, 1971) may not be applicable in grouted base
plate details. While it was determined that the shear capacity of bolts for grouted plates is
reduced as compared with ungrouted plates (partly due to the localized damage of the
grout), no design recommendations were given. Analytical design aids of bolts subjected
to tension plus shear loading are presented by Scacco (1992), Fisher (1981) and Shipp &
Haninger (1983). More recently, Nakashima (1998) examined the mechanical properties
of steel column base anchor bolts under combined tension and shear. In addition to
10
examining the effects of anchor rod threads, Nakashima determined that, due to grout
separation, the ultimate capacity of anchor bolts placed through a grout layer is lower as
compared with typical structural bolts. Kawano et al. (2003) experimentally studied
anchor bolts in exposed steel column bases, focusing on bolts which are subjected to
tensile and shearing forces with relatively short concrete edge distances.
Current experimental and theoretical research on column bases loaded with a

combination of shear and tension forces by Gresnigt et al. (2008) specifically investigates
anchor rod bearing. Gresnigt et al.’s results are consistent with past studies which show
that the shear strength of anchor bolts is considerably lower than the shear strength of
bolts in bolted connections between steel plates. Two analytical capacity models are
reported to describe the anchor rod load-deformation behavior. One model gives very
conservative results, especially with “large” tensile axial forces and/or “large” grout
thicknesses. Another proposed model, which does not take into account bending
moments in the bolts, gives results that are consistent with test results.
In summary, studies on the behavior of anchor rods have focused mainly on concrete
failure modes, and on the axial-shear interaction of the anchor bolts. The behavior of
anchor rods in grouted base plates has received limited attention by way of experimental
studies. Moreover, additional studies are required to appropriately characterize the effects
of bending on the capacity of anchor rods.
2.3.3 Shear Key Bearing

To resist moderate to high column base shear loads, such as induced by seismic loading
in low- to mid-rise structures, one or multiple shear lugs (also known as a shear key) may
be attached to the base plate. The shear key is often provided in the form of a plate
welded to the underside of the base plate. In some cases, a stub W-section is used to
provide the required lug strength. The shear lug detail requires additional welding and the
need for a grout pocket in the concrete foundation, which may lead to more costly and
difficult construction.
11
Guidelines and examples for shear key design are provided in the AISC Steel Design
Guide 1. Transfer of shear through shear key bearing may be accompanied by three
failure modes including: (1) failure of the grout/concrete directly in front of the load path
of the shear key, (2) failure of the steel shear key itself, including the welded connection
to the base plate and (3) distortion (e.g. bending) of the base plate due the loading of the
connected shear key.
With respect to issue (1) described above, three typical failure modes of the
concrete/grout are associated with the shear key mechanism: (a) concrete/grout bearing
failure directly in front of the shear key bearing surface load path, (b) concrete shear-
friction failure and (c) concrete free edge side blowout failure. Failure modes (a) and (b)
are common for foundations with large areas surrounding the shear key while failure
mode (c) is common for foundations with short edge distances from the shear key such as
in pier footings. Provisions for failure modes (a) and (b) are provided in ACI 349 and are
based on tests by Rotz & Reifschneider (1989). Regarding concrete side blowout, a
literature search did not reveal any experimental data for base plate shear lugs. ACI 349
provides two methods to predict the blowout shear capacity of concrete embedments
toward a free edge: the 45 degree cone method and the concrete capacity design (CCD)
method.
The 45 degree cone method for concrete embedments was developed by the ACI
Committee 349 (ACI 349, 1985). As per this method, the concrete blowout capacity of a
fastening/embedment subjected to tension or shear is calculated by assuming a constant
tensile stress acting on the projected area of a failure cone, assuming the inclination
between the failure surface and the concrete surface as 45 degrees. This method was
adapted by ACI 349 (ACI 349, 2006) for the shear capacity of shear lugs loaded towards
a free edge, assuming a 45 degree failure plane from the bearing edges of the shear key.
However, no experimental tests were conducted to verify this method in the context of
shear keys attached to base plates.
12
The CCD method (Fuchs et al., 1995) was developed in response to the limitations of the
45 degree cone method and is currently the standard design method for concrete anchors
under shear and tensile loading (ACI 349, 2006). The CCD method is based on the
principle that for large concrete members, the blowout strength is associated with fracture
initiation, rather than the development of concrete tensile strength over a large failure
surface, such as assumed by the 45 degree cone method. Thus, for large foundation sizes,
the 45 degree cone method may overestimate the available strength. To address this issue,
the CCD method (Fuchs et al., 1995) applies the fracture mechanics theory of Bažant
(1984) to concrete anchors, by explicitly including a size effect, wherein the failure stress
of concrete is assumed to be an inverse function of the specimen/foundation size. While
the theory is presented for concrete anchors, no experimental data is available to
verify/examine the CCD method in the context of shear keys in base plates.
With respect to (2) and (3) described earlier, the design of shear lugs itself is presented in
Fisher (1981), Goldman (1983), Tronzo (1984) and more recently in the AISC Steel
Design Guide 1 (Fisher & Kloiber, 2006). The AISC Design Guide recommends the lug
be treated as a cantilever beam with a uniform bearing stress on the face of the lug, from
which the lug thickness and weld size can be determined. The design guide also mentions
that bending in the base plate may be an issue for large shear loads and/or column weak
axis loading. However, a literature search did not reveal any experimental investigations
addressing this issue.
In summary, studies investigating the response of base plate shear key bearing are
extremely limited. Owing to the unavailability of experimental data for shear lugs, the
aforementioned research synthesis by Grauvilardell et al. (2005) states that research on
this topic is priority.
2.4 OBJECTIVES OF CURRENT RESEARCH

As discussed in the preceding sections, several topics regarding shear transfer in base
plates are unresolved and may be addressed through further experimentation and analysis,
especially in the context of large scale prototypes. Based on consultation with the AISC
13
oversight committee and consideration of experimental budgets, the main objectives of

the current study are to examine three mechanisms (surface friction, anchor rod bearing
and shear key bearing) commonly used to resist shear in exposed column base plate
connections. While several failure modes are possible for each of these mechanisms, the
specific issues addressed by this study are now summarized –
1. For surface friction, the major objective to is characterize the coefficient of

friction between steel and grout, and more importantly, between the steel and
grout interface when leveling shim stacks are used.
2. For anchor rod bearing, the major objective is to characterize the strength of the
anchor rod itself under the combination of axial tension, shear and flexure. Three
issues are important in this context. First, the effective bending length and
bending shape of the anchor rod must be characterized to determine the flexural
stresses in the rod. Second, the effects of interaction/contact between the anchor
rods and base plate on the strength of the anchor rods must be examined. Third,
geometric second order effects due to large displacement of the rods should be
considered. Concrete failure modes are not explicitly investigated.
3. For shear key bearing, the major objective is to investigate the strength capacity
associated with the side blowout of a shear key embedded near a free edge in an
unreinforced concrete footing. Other failure modes, such as shear key yielding or
base plate failure due to the forces imposed by the shear key are not considered.
In addition to providing experimental data and analysis for each of these shear
mechanisms, a key objective of the current study is to verify existing design methods and
to provide improved design guidelines for shear transfer. Chapter 3 describes the results
of the experimental program in detail, including an analysis of the test observations.
14
Chapter 3: Experimental Program and Test Results
Chapter 3
Experimental Program and Test Results
3.1 INTRODUCTION
Observations from seven large scale tests are reported to investigate shear transfer in
exposed column base plates. Three mechanisms, commonly used in current structural
design practice, are examined: surface friction, anchor rod bearing and shear key bearing.
While similar tests have been conducted previously in broader contexts, the tests
described in this chapter focus on shear transfer mechanisms from the specific
perspective of exposed column base plates. For example, the surface friction tests
investigate the effect of shim stacks typically used in base plates for column erection (as
opposed to only considering pure steel on grout friction). Similarly, tests involving the
anchor rods investigate the effect of anchor rod bending, rather than only considering the
interaction of axial and shear stresses. The shear key tests provide entirely new
experimental data on concrete blowout behavior in an unreinforced footing.
The large scale experiments, along with the ancillary tests (also described in this chapter)
provide data which can be used to evaluate current design guidelines and capacity
predictions, such as those outlined in the AISC Steel Design Guide Series 1 (Fisher &
Kloiber, 2006). This chapter describes the experimental program in detail by providing an
overview of the large scale test setup, testing procedures and test results. While this
chapter presents key data and representative response plots associated with the discussion
of experimental observations, detailed data for all experiments is provided in Appendix
B. An analysis of observations and a brief assessment of current design guidelines are
provided for each shear transfer mechanism investigated. This chapter also presents a
summary of the ancillary tests, including anchor rod tension tests, standard concrete
cylinder compression tests and grout cylinder compression tests. An inventory of the
ancillary tests, along with material descriptions, is given Appendix A.
15
3.2. ANCILLARY TESTS

Three types of ancillary tests were conducted to characterize the materials used in the
base plate tests. These include (1) anchor rod tension tests, (2) standard concrete cylinder
compression tests and (3) grout cylinder compression tests. Tables 3.1-3.3 summarize key
results obtained from the ancillary tests and Appendix A includes a detailed inventory of
the ancillary test data.
3.1.1 Anchor Rod Tension Tests

Tension tests were conducted on samples taken from the same batch of ASTM F1554
Grade 55 anchor rods used for the base plate tests. Four unthreaded rod specimens (two
for each rod size used in the large scale tests, 3/4” and 1-1/4” diameter) were tested
quasi-statically under monotonic tension to determine anchor rod material properties.
Two uniaxial strain gages were attached to the rods to identify initial yielding behavior.
As summarized in Table 3.1, the mean ultimate strength Fu,rod is 96.4 ksi for the 3/4”
B B
diameter rod and 75.0 ksi for the 1-1/4” diameter rod. These values are approximately
within the ASTM specified ultimate strength range of 75-95 ksi (AISC, 2005). The
average yield strength Fy,rod (as determined by the 0.2% offset method) is 66.8 ksi for the
B B
3/4” diameter rod and 54.4 ksi for the 1-1/4” diameter rod. The yield strength of the
smaller 3/4” diameter rod is approximately 21% greater than the minimum specified
strength of 55 ksi, whereas the yield strength of the 1-1/4” diameter rod is about equal to
the specified strength. Complete results of the rod tension tests are provided in Appendix
A.
3.1.2 Standard Concrete Test Cylinders

Four pedestals representing the concrete foundation for the base plate tests (see Figure
3.1), measuring 4 feet by 7 feet in plan area and 32.5 inches in height, were cast using
commercial ready-mixed concrete specified with a 4,000 psi twenty-eight day
compressive strength and 4.0” slump. The concrete was delivered in two trucks and each
batch had a measured slump of 3.5”. Prior to testing, all concrete pedestal specimens
were air cured well beyond 28 days (i.e. between 3-5 months). At this age, similar
16
concrete is approximately 13% stronger than the measured twenty-eight day compressive
strength (ACI 209, 2006). Note that for the purpose of analysis, the twenty-eight day
compressive strength is considered. A total of 12 concrete compression test cylinder
specimens (three cylinders sampled from the concrete used for each pedestal) were
collected (as per ASTM C31, 2008) and tested after twenty-eight days of curing (as per
ASTM C39, 2005). The average ultimate compressive strength f c′ of all test cylinders is
4,760 psi (standard deviation = 260 psi), about 19% greater than specified (4,000 psi).
The average ultimate compressive strength of the concrete from each pedestal is shown in
Table 3.2, along with other information such as cylinder density. Detailed information
regarding the concrete, including the test results of each individual compression test, is
archived in Appendix A.
3.1.3 Grout Test Cylinders

General construction, mineral-aggregate non-shrink grout was installed between the steel
base plate and concrete pedestal for all large scale tests. Tests were conducted to examine
the compressive strength of the grout used for each base plate test. The grout compressive
strength is especially important for the large scale tests involving the shear key, where the
failure surface encompasses both concrete and grout and grout bearing failure is a
possible failure mode. Twenty-eight grout cylinders (four samples from the grout used
for each large scale test), measuring 6” tall and 3” in diameter, were tested in
compression at a loading rate of approximately 3.5 kips/second. The standard ASTM
method for testing grout strength (ASTM C109, 2007) was not used. This method, also
known as a “grout cube test”, is applicable for masonry grout applications. Except for the
smaller cylinder size, the grout was tested and collected using the same methods as for
standard concrete cylinders (i.e. ASTM C31 and ASTM C39). For logistical (laboratory
schedule) and construction reasons, the grout pad from each large scale test had different
cure times and water content. This variation in the cure time and water content was
replicated for each grout test cylinders. The average compressive strengths of the grout is
presented in Table 3.3, along with other information such as the corresponding large
scale test number, curing time, water content, and density. The average compressive
strength of the grout ranged between 5,780 and 7,210 psi. Standard practice typically
17
requires the grout to have a compressive strength at least twice as large as the
compressive strength of the supporting concrete (i.e. 8,000 - 10,000 psi grout in the case
of the concrete used in the large scale experiments). Due to testing schedule constraints,
the grout was not cured to full strength. Detailed information regarding the grout,
including the test results of each individual compression test and product data from the
manufacturer, is archived in Appendix A.
3.3 LARGE SCALE EXPERIMENTS: TEST SETUP AND PREPARATION

Table 3.4 presents the test matrix for the base plate tests. The table includes key
information about each experiment, including information regarding the shear mechanism
investigated and the applied loading. One concrete pedestal (which represents the
foundation/footing) was used for all the friction tests (Tests #1, #2 and #3), as well as one
anchor rod test (Test #4). A new pedestal was used for each of the other three tests (Tests
#5, #6 and #7). Figure 3.1 shows a representative photograph of the concrete pedestal.
The three mechanisms investigated for the base plate tests, indicated in Table 3.4,
represent popular design alternatives for shear transfer in exposed base plates and are
featured in the AISC Steel Design Guide 1 (Fisher & Kloiber, 2006). Moreover, these
mechanisms were enumerated as being important by the AISC (AISC RFP 6807, 2006) in
response to which this study was initiated. Referring to Table 3.4, surface friction is
examined under cyclic shear and various compressive axial load levels. To reflect
common construction procedures, two friction tests (Tests #1 and #2) included steel shim
stacks beneath the base plate. For comparison, a third friction test (Test #3) investigated a
steel on grout interface without the shim stacks. Two tests (Test #4 and #5) were
conducted to investigate the response of anchor rods in a grouted base plate connection
under combined shear and tensile loading. Two differently sized anchor rods (3/4” and
1-1/4” diameter), with a welded plate washer detail, were investigated. Two other tests
(Tests #6 and #7) focused on a shear key bearing mechanism. Recall that a literature
review did not reveal any prior tests which evaluated shear keys loaded toward a free
edge, even though this detail is often suggested for exposed base plates with high shear
18
forces. Two shear key embedment depths, 5.5 inches and 3.0 inches below the concrete
surface, were tested under monotonic shear with small axial compression force.
3.3.1 Test Setup

The base plate tests were conducted at the UC Berkeley Network for Earthquake
Engineering Simulation (NEES) Structures Laboratory in Richmond, California. Figure
3.2a schematically illustrates the test setup and a photograph is provided in Figure 3.2b.
Referring to Figure 3.2a, a large steel beam-column loading frame provided a rigid load
path for both compressive and tensile axial loading. The loading frame was braced
laterally onto a concrete reaction wall. Horizontal (i.e. lateral) loads were applied via two
steel assemblies bolted directly onto the base plate (see Figure 3.3), thus applying direct
shear loading with negligible moment at the base connection. Two MTS series 244 150-
kip dynamic actuators provided the vertical (axial) loading while two MTS series 244
220-kip dynamic actuators provided the horizontal (shear) loading. The test assembly,
including the base plate, was designed to remain undamaged during testing and was
reused for all seven tests. However, the base plate was modified for certain tests,
including the welding (and removal) of plate washers (for Tests #4 and #5) and the
welding (and shortening) of the shear key (for Tests #6 and #7). The total self weight of
the set-up, determined by the vertical actuator load cells, was approximately 17.2 kips.
This weight was subtracted from the axial loads recorded from the vertical actuators to
determine the net axial load in the base plate connection.
3.3.1.1 Test Setup for the Friction Tests: Tests #1, #2 and #3
For Tests #1 and #2, steel shim stacks (see Figure 3.4) were used to support the base plate
on the concrete pedestal during column erection. While other erection procedures were
considered, the shim stack method is commonly used in current practice (Fisher &
Kloiber, 2006). The shims were thermally cut from steel bar stock, each approximately 4
inches by 2 inches in area and 1/2 inch thick. Each shim stack consisted of two shims,
resulting in a one inch clear distance between the surface of the concrete and the base
plate, thereby providing a one inch grout pad thickness. The shims were heavily oxidized
and had a slightly rough surface, especially around the shim edges. Deposits resulting
19
from thermal cutting were chipped away from the surface of the shims but the shims were
not de-burred nor treated in any other way. Thus, the shims were fabricated to reflect
standard construction practice. As shown in Figure 3.5, different shim stack positioning
was used for Test #1 and #2. Test #3, designed to examine friction between steel and
grout, did not feature shim stacks and the column was lowered by the vertical actuators to
provide a grout pad thickness of one inch.
Anchor rods were absent for all friction tests in order to prevent unwanted strength
contribution from rod bearing. Grout was installed by pouring a flowable mix through
one anchor rod hole in the base plate until it flowed out through the other three anchor
rod holes. A short tube, placed over one anchor rod hole, was used to provide head
pressure to distribute the grout. The concrete pedestal surface was moistened to ensure an
easy flow of the grout. Post-test visual inspection of the grout pad confirmed that the
grout was evenly distributed and contained few air voids (see Figure 3.6). For all friction
tests, the grout was contained within an approximate 26” by 26” foam dam, thus
providing a 4.70 square foot grout pad area (see Figs. 3.5 and 3.6). Excess grout was
removed from the base plate anchor rod holes prior to curing. Compressive loading, equal
to the test setup assembly weight (17.2 kips; 25 psi), was maintained in the connection
during curing to preserve the steel to grout bond.
Since frictional response is sensitive to surface properties, it is relevant to discuss the

surface condition of the base plate. Before the first friction test, the surface condition of
the ASTM A529 Grade 50 base plate was typical of freshly fabricated steel. The plate
contained some mill scale and did not have any visible signs of oxidation. To reflect field
conditions, the surface was not treated (sanded nor polished, etc.) in any way. After the
first friction test, the plate sustained buffing damage from the grout and small patches of
scouring damage due to the steel shim stacks, but no other damage was visible. Prior to
the next two friction tests, residual grout was removed from the plate and the surface was
wiped clean with a damp cloth. As shown previously in Fig. 3.5, different shim stack
positioning was used for Tests #1 and #2 such that the scoured plate surface from Test #1
did not coincide with the shim stack locations of Test #2.
20
3.3.1.2 Test Setup for the Anchor Rod Tests: Tests #4 and #5
Figure 3.7 shows a schematic of the rebar details of the concrete pedestals used for the
anchor rod tests (Test #4 and #5). As indicated in the figure, the concrete pedestals
include specifically designed rebar details to ensure failure of the anchor rod itself, rather
than failure within the concrete. Four anchor rods were used and the rod layout was
approximately 24” inches square for all tests (see Figure 3.8). The anchor rod holes in the
two inch thick base plate were 2-1/16” in diameter, the recommended size for 1-1/4”
diameter anchor rods stated in the AISC Steel Design Guide 1 (Fisher & Kloiber, 2006).
Note that since the same base plate was reused for both rod sizes, the rod hole size was
larger than the recommended hole diameter of 1-5/16” for the 3/4” diameter anchor rods.
As recommended by the AISC Steel Design Guide 1 (Fisher & Kloiber, 2006), plate
washers, provided for the anchor rods, were fillet welded to the base plate to reduce the
amount of slip before the shear loading was transferred to all rods. Machined square plate
washers, measuring 2.5” x 2.5” x 1/4” with an internal diameter of 0.8 inch (≈ rod
diameter + 1/16”), were used for the 3/4” anchor rods of Test #4 and plate washers,
measuring 3.5” x 3.5” x 1/2”, with holes approximately 1.3 inches in diameter (≈ rod
diameter + 1/8”), were fabricated from thermally cut plate stock for the 1-1/4” diameter
rods of Test #5. The sizes and hole diameters of these plate washers reflect
recommendations from the AISC Steel Design Guide 1 (Fisher & Kloiber, 2006).
Additional washers were placed (un-welded) over the rods on top of the plate washers to
prevent dishing of the welded plate washers due to the large tension forces in the rods. A
photograph of this detail is shown in Figure 3.9. Nuts were hand tightened snug with an
additional 1/8 turn. For both tests, a stiff grout mix was placed on the concrete pedestal
and the base plate was lowered on to it. The grout was compacted to an approximate 1-
1/4” thickness for Test #4 and an approximate 1” thickness for Test #5. Excess grout was
removed from the anchor rod holes prior to curing. The threads of the anchor rods
extended approximately 2 inches below the surface of the concrete to ensure failure in the
threaded region.
21
3.3.1.3 Test Setup for the Shear Key Tests: Tests #6 and #7
For Tests #6 and #7, an I-shaped shear key, fabricated from plate stock, was welded (with
1.5” fillet welds) to the center of the base plate (see Figure 3.10 and 3.11). The shear key
was designed to remain undamaged during testing. For Test #6, the shear key was 7
inches long. After this test, the shear key was thermally cut to reduce its length to 4.5
inches for Test #7. The bearing width of the shear key for both tests was 6”. Referring to
Fig. 3.1, grout pockets were provided in the concrete pedestals, measuring 9” x 9” in
plan, 7” deep for Test #6 and 4” deep for Test #7. The distance from the shear key
bearing surface to the edge of the concrete pedestal was about 20.25 inches. Grout was
installed by pouring a highly flowable mix through two grout holes in the base plate.
Shims were used for both shear key tests to provide a grout thickness of 1.5”, resulting in
an approximate shear key embedment depth (below the surface of the concrete) of 5.5”
for Test #6 and 3.0” for Test #7. Anchor rods (3/4” in diameter) were cast in the pedestal
to mimic the reinforcing aspect of the anchor rods, but the rods were not bolted to the
base plate in order to isolate the strength of the shear key. Figure 3.12 shows the
reinforcement layout of the pedestals used for the shear key tests. It is important to note
that the anticipated blowout area (i.e. failure region) of the concrete pedestal was free of
any reinforcement or other obstructions (with the exception of the anchor rods), to ensure
that failure was obtained in the concrete only. For Tests #6 and #7, the concrete pedestals
were attached to the laboratory strong floor by tension rods that terminated below the
anticipated blow out area (refer Fig. 3.12), so as not to interfere with the concrete
blowout failure mechanism.
3.3.2 Loading Protocol

Table 3.5 indicates the loadings applied in the base plate tests. Detailed loading plots for
each test are presented in Appendix B. As indicated in Table 3.5, the friction tests
involved the application of different axial load levels, followed by cyclic lateral
displacement based loading as indicated in Fig. 3.13. As shown in the figure, the cyclic
loading history consists of sets of increasing reversed lateral slip cycles (three cycles per
set), with amplitudes of 0.1”, 0.2”, 0.4”, 0.8” and 1.0”. It is relevant to note that while
several loading protocols are available for structural component testing, they are typically
22
expressed in terms of interstory drifts with the intent of testing deformation sensitive
components. Similar protocols are not readily available for “stiff” mechanisms such as
the shear transfer examined in these tests. Thus, the lateral deformation histories applied
in this study are adaptations of other such protocols (i.e. Krawinkler et al., 2000) to
reflect deformation demands that may be expected during shear transfer. For Test #1,
constant axial loads levels were applied in increasing orders of magnitude such that the
steel-grout bond was broken at the lowest axial load level. Three axial loads were applied
(43, 112 and 261 kips), and the cyclic lateral loading history (see Fig. 3.13) was applied
at each stage. These axial loads correspond to concrete foundation bearing stresses of
about 64, 166, and 386 psi, which is approximately two to ten percent of the nominal
ultimate compressive strength (4,000 psi) of the concrete. For Test #2 and #3, the axial
loads were applied in reverse (i.e. decreasing) order, thereby breaking the grout bond at
the highest load level.
Referring to Table 3.5, similar loading was applied for the anchor rod tests (Tests #4 and
#5), except that the constant axial load was tensile. Table 3.5 indicates the level of tensile
load applied to each of the anchor rod bearing tests. Considering equal load distribution
between the rods, the axial load levels were about 31% (Test #4) and 39% (Test #5) of
the ultimate tensile capacities of the anchor rods (based on the measured material
properties).
For the shear key tests, a relatively small level of axial compressive load (between 0-27
kips for Test #6 and 15 kips for Test #7) was applied to prevent liftoff of the column from
the concrete pedestal. For both tests, lateral (shear) deformations were applied
monotonically in one direction until failure. Subsequently, the loading was reversed and
applied monotonically in the opposite direction until failure was achieved.
3.3.3 Instrumentation
Figure 3.14 illustrates the typical instrumentation layout for the base plate tests. In
addition to the loads and displacements monitored by the four actuators themselves,
instrumentation was installed on the specimens and loading frame to measure
23
displacements and strains. These instruments included position transducers attached to

the base plate and concrete pedestal to measure base plate and concrete pedestal motion
in three dimensions, thus enabling an accurate monitoring of the relative slip between the
base plate, concrete pedestal and laboratory strong floor. Similarly, displacement
transducers were installed to monitor unwanted out-of-plane motion of the test loading
frame. Strain gages were attached to the testing rig to monitor deformations within the
test setup. For Tests #4 and #5, two uniaxial strain gages were attached to the anchor rods
(at two inches below the upper threads; about five inches below the surface of the
concrete) to measure rod axial strain. Strain gages were also attached to the shear key for
Test #6 (see Figure 3.11). For Tests #4 and #5, embeddable concrete strain gages were
installed to monitor concrete strains near the anchor rods. Embeddable concrete strain
gages were also installed, in the vicinity of the concrete failure region, for Test #6 and #7.
Prior to and following testing, the geometry of the concrete pedestal for Test #6 was
measured by a high definition laser scanner to generate three dimensional plots of the
failure surface (see Figure B.14 in Appendix B).
3.4 LARGE SCALE TEST RESULTS

This section presents key observations from the base plate tests. In addition, the test
results are analyzed and an evaluation of current design guidelines is given. Detailed
experimental data is archived in Appendix B. A summary of results and a discussion of
design implications are presented in Chapter 4. The discussion of the test observations is
divided into three sections, each focusing on one specific shear transfer mechanism, i.e.
surface friction, anchor rod bearing and shear key bearing.

As discussed in the previous sections, Tests #1-3 involved the application of three
compressive axial loads levels (43, 112, and 261 kips; with corresponding bearing
stresses of about 64, 166, and 386 psi) between the base plate and grout interface. For
each constant compressive load level, a lateral displacement-based load protocol
(illustrated in Figure 3.13) was applied. Figure 3.15 shows the representative response of
the friction tests (shown here for Test #3). This figure shows a large initial lateral
24
(frictional) force due to steel-grout bonding followed by a square shaped hysteretic

response indicative of Coulomb friction behavior. Similar behavior is observed for all
friction tests (refer Appendix B). Note that all slip occurs between the grout and the steel
base plate, rather than the grout and the concrete pedestal.
Figure 3.16 shows the condition of the grout, base plate and shim stacks after Test #1. In
addition to mild abrasion marks observed over the entire base plate, moderate scouring
damage to the base plate was localized at the locations of the shim stacks. For Test #1,
only two shim stacks exhibited significant scuffing damage, indicating that not all shims
came into complete contact with the base plate during column erection. On the other
hand, Test #2 showed approximately equal scuffing damage to all shims, suggesting that
all were in complete contact with the base plate. Aside from mild spalling of the extreme
perimeter of the grout pad, no damage to the grout (other than abrasion by the steel base
plate) was observed for all surface friction tests. Moreover, for all friction tests, no
damage to the concrete was observed, and the grout retained its bond to the concrete.
Figure 3.17 plots the frictional force versus cumulative slip of the base plate for Test #2
(similar graphs for the other friction tests are presented in Appendix B). The quantities
are plotted over the entire duration of Test #2, illustrating the large jumps in frictional
force when the normal (axial) load is changed. Figure 3.18 plots the corresponding
frictional force normalized by the normal (axial) load (thus providing an effective
coefficient of friction) for each friction test. As observed from Figure 3.18, the lateral
resistance (i.e. the effective coefficient of friction), varies significantly during the
duration of each normal load level. In addition, for each experiment (refer Appendix B
and Figure 3.18), the initial force required to overcome the steel-grout bond resistance is
significantly higher (approximately twice) than the frictional force remaining after the
bond is broken. Thus, characterization of the coefficient of friction is somewhat
subjective. Four key observations are made to develop a strategy to characterize the
coefficient of friction in a consistent and conservative manner –
25
1. For all friction tests (Tests #1, #2 and #3), the initial resistance corresponding to steel-
grout bond breakage is significantly higher than the frictional force following bond
breakage. This bond resistance arises from chemical bonding (i.e. constitutive
behavior) that is distinct from sliding friction resistance. Moreover, this bond
resistance may not be available in field conditions where the bond could be broken
during construction or over the life of the building due to overloads. Thus, shear
resistance corresponding to the bond force is not considered in the characterization of
the coefficient of friction because it may provide an unconservative and unrealistic
estimate.
2. For both the experiments with shim stacks, i.e. Tests #1 and #2, the effective
coefficient of friction follows a consistent pattern of evolution. After the initial bond
breakage, the shear resistance remains relatively constant during the initial
displacement cycles. Subsequently, the shear resistance increases, which may be
attributed to the shim stacks gouging into the underside of the base plate (see Figure
3.16). As this mechanism occurs only after significant cumulative lateral slip, this
increase in the friction coefficient may not be suitable for design. Accordingly,
coefficient values corresponding to shim stack gouging are not considered in the
analysis.
3. The minimum coefficient of friction values extracted from Tests #1 and #2 are shown
in Figure 3.18a and Figure 3.18b, respectively. Following the observations discussed
above, these values, which are considered in the determination of the friction
coefficient (1) are the lowest frictional forces observed for each axial load level and
(2) do not include the increased frictional resistance due to shim stack gouging nor
steel-grout adhesion. For Test #2, an increase in resistance from steel shim gouging is
evident during the first level (261 kips) of axial compressive loading. Thus, only one
friction coefficient value is extracted from Test #2 (see Fig. 3.18b). For Test #1, two
values can be extracted since the increase in frictional resistance is delayed until the
second level (112 kips) of axial loading. This delay suggests that steel shim gouging
26
occurs at lower levels of cumulative slip for higher levels of compressive axial
loading.
4. Figure 3.18c shows the values extracted from Test #3 for the determination of the
friction coefficient. These values do not consider the initial larger force corresponding
to bond breakage and include the lowest frictional force points observed for each of
the three different gravity load levels. Since Test #3 does not feature shim stacks, a
significant increase in frictional resistance due to steel shim gouging is not observed .
Slight dips in the effective coefficient of friction observed subsequent each change in
axial load level (as seen in Figure 3.18a-3.18c) occur during cyclic displacement
amplitudes of 0.1” (recall Figure 3.13) under which the slip velocity is slower (4.2 x
10-3 inches/sec) than the rest of the loading (1.4 x 10-2 inches/sec). Thus, there is a
P P P P
possibility that frictional resistance is sensitive to slip velocity. Prior friction tests of
non-metallic materials on steel (Fenz, 2002) show that frictional resistance decreases
as the slip velocity tends to zero.
Figure 3.19 plots the frictional force values (extracted as described in the preceding
discussion) versus the normal (axial) load values collected from the friction tests. These
values are also listed in Table 3.6. Figure 3.19 also shows a linear regression line (with
the intercept set to zero) for the data set from Test #1 and #2, from which the coefficient
of friction is determined as 0.46 (for a base plate on a grout pad with shim stacks). A
similar regression line for Test #3 results in a coefficient of 0.45 (for base plate on a pure
grout pad). Both these regression lines have a high R 2 value (greater than 0.98)
indicating a good fit and confirming the expected linear frictional response. Appendix B
presents comprehensive data regarding the surface friction tests (i.e. Tests #1, #2 and #3).
A review of this data indicates that the values presented in Figure 3.19 and Table 3.6 for
the determination of the friction coefficient are the most conservative, whereas including
additional data (i.e. considering shim stack gouging and steel-grout bond adhesion)
shows significant deviation from the expected linear relationship for friction.
27
The AISC Steel Design Guide 1 (Fisher & Kloiber, 2006) lists the coefficient of friction
as 0.55 for steel on grout. This is adapted from ACI 349-85 which states, “μ = 0.55 for
grouted conditions with the contact plane between grout and as-rolled steel exterior to
the concrete surface”. This value is based on a reported value of 0.53 from Cannon et al.
(1975). Recent tests by Nagae et al. (2006), report a coefficient of friction value between
steel and mortar as 0.52. The value determined by the current study (0.45) is lower than
these previously reported values. A study investigating the coefficient of friction between
a steel base plate and concrete (Cook & Klingner, 1991) reports an average coefficient of
0.43. It is important to note that the friction coefficient may be highly sensitive to the
condition of the steel surface, as well as grout properties and construction procedures. For
example, under the current study, the grout was allowed to cure for only about one week
due to laboratory scheduling requirements. Moreover, the coefficient friction may be
highly sensitive to the presence of mill scale on the steel base plate. A series of friction
tests between steel and concrete by Baltay and Gjelsvik (1990) indicates that the
coefficient of friction for a mill scale steel surface is less than that for a machined (i.e.
polished) surface for bearing stress levels below 10,000 psi. The report describes that the
mill scale is harder than the steel and is therefore not penetrated by the concrete/grout
particles at lower bearing stress levels, resulting in a lower coefficient of friction. The
conservative value μ = 0.45 for the coefficient of friction between a steel base plate with
mill scale and grout pad (with or without steel shim stacks) is recommended based on
experimental data from the current study.

Tests #4 and #5 investigate base plate shear resistance via four anchor rods under a
combination of imposed axial tensile loads and shear/flexural loading. As described
previously, the connection includes welded plate washers and grout between the base
plate and concrete pedestal. The main difference between Tests #4 and #5 is the nominal
anchor rod size; Test #4 features 3/4” diameter anchor rods while Test #5 features 1-1/4”
diameter rods. All rods were ASTM F1554 Grade 55 ksi steel and were installed such
that the rod threads ran below the surface of the concrete pedestal; i.e. the threads were
included in the shear plane. Due to the threads, the minimum root diameter for the 3/4”
28
bolts was 0.64" and 1.10” for the 1-1/4” bolts. Table 3.5 indicates the tensile axial load
applied to each anchor rod test. Compared to the measured strengths presented previously
in Table 3.1, the applied axial tension load was approximately 31% of the ultimate load
(of all four rods) for Test #4 and 39% for Test #5. The axial load was held constant and
was followed by the application of cyclic lateral displacements (as described previously
and illustrated in Figure 3.13).
Qualitatively, both anchor rod experiments (Tests #4 and #5) followed a similar
progression of events. Visual inspection indicated that the application of tensile axial
loading instantly broke the bond between the steel base plate and the grout pad. Figure
3.20 plots the vertical displacement (averaged from four locations on the base plate)
versus the initial axial load for both tests. Under initial tensile axial loading, the base
plate displaced vertically upward 0.035” for Test #4 and 0.07” for Test #5. As observed
through anchor rod strain data, the axial load distribution during initial vertical loading
was not uniform for all rods, especially for Test #4 (see Figure 3.21). However, no rods
yielded during the initial base plate uplift. Subsequent to this, the axial loading produced
no discernible indications of damage, except that the gap between the base plate and the
grout pad increased as continuing lateral deformations were applied. Figure 3.22 plots the
base plate vertical displacement (averaged from four locations on the base plate) during
the course of the cyclic lateral loading (expressed as cumulative lateral displacement).
Prior to failure, the base plate displaced vertically approximately 0.35” for Test #4 and
0.50” for Test #5, illustrating axial elongation of the anchor rods caused by the cyclic
lateral displacement of the base plate under constant axial tension.
Figures 3.23 and 3.24 plot lateral load versus lateral base plate displacement for Tests #4
and #5, respectively. An analysis of these plots is provided in a later section. Both tests
were concluded when one rod fractured during a cyclic displacement excursion (see
Figure 3.25). For Test #4 (3/4” diameter rod), rod fracture occurred during the 4th P P
amplitude of the 0.8” cycle and for Test #5 (1-1/4” diameter rod) during the 1st excursion
P P
of the 1.0” cycle. Table 3.7 summarizes the experimental results for both anchor rod tests,
including the peak lateral load observed for each slip direction. In addition, Table 3.7 lists
29
predicted anchor rod strengths (to be discussed later). The fracture points are indicated on
the load deformation curves in Figs. 3.23 and 3.24. Fig. 3.26 overlays a photograph of the
fractured anchor rod from both tests with the locations of the base plate, plate washers,
grout and concrete.
Inspections made after Test #4 (3/4” rods) reveals damage occurred to the grout pad in
the form of localized spalling around the anchor rods. Furthermore, at two anchor rods
locations, the grout cracked and completely separated from the concrete pedestal (see
Figure 3.27a). However, no damage to the concrete pedestal was observed for Test #4.
An anchor rod from Test #4 fractured at about 1/4” from the top of the concrete surface
(i.e. within the grout pad – refer overlay schematic in Fig. 3.26a). Test #5 showed
evidence of slight concrete damage (local spalling cones around the perimeter of each
rod, generally 2” to 4” in diameter and 1” to 1.25” deep – see Fig. 3.27b) and extensive
grout damage (see Fig. 3.27c). For Test #5, the anchor rod fractured within the concrete
pedestal, about one inch below the surface of the undamaged concrete (refer overlay
schematic in Fig. 3.26b). After the completion of Test #5, two of the anchor rods were
excavated from the concrete and exhibited significant residual deformations up to about
one inch below the surface of the damaged concrete (i.e. where the rod threads
terminated). For both tests, damage to the grout was so extensive that pieces of the grout
pad completely separated from the concrete pedestal. Cracking of the grout pad was
observed during the 0.2” amplitude displacement cycle for both anchor rod tests. Note
that the grout was installed up to the edge of the steel base plate such that the edge
distance between the rods and grout pad was approximately five inches.
Referring to the load-deformation Figures 3.23 and 3.24, a complex hysteretic response,
controlled by several phenomena, is observed. The following explanation of the response
presented in this section is based on visual observations during testing, post-test
inspection of the deformed anchor rods and damaged grout, as well as theoretical
analyses of the anchor rods. Appendices C and D summarize the theoretical analyses of
the anchor rod tests; Appendix C provides an examination of current design methods
while Appendix D provides a detailed analysis incorporating geometric and material
30
nonlinearities. Several observations may be made from Figures 3.23 and 3.24, the
observed response of the specimens and the associated analyses (Appendix D). These are
now summarized –
1. The large initial lateral force observed in Test #4 (see Figure 3.23) is attributed to
stubs of grout (approximately 0.75” tall) which were not completely removed
from the four base plate anchor rod holes. The highly confined grout stubs
constrained the movement of the anchor rods or the base plate, thus requiring a
relatively high load to break. This highly brittle mode of failure (and its
corresponding peak load) is neglected in the analysis since it corresponds to a
certain construction procedure which can not be relied on in practice.
2. For the anchor rod tests, it is challenging to characterize a peak load since the rod
response is controlled by several phenomena, including the yielding of the anchor
rods in bending followed by an increase in strength due to second order geometric
effects as the bolted end of the yielded rods displaces laterally, thereby resulting
in a post-yield increase in lateral load. Figure 3.28 illustrates this effect, in which
the axial force in the rods increase as they deform, resulting in increased lateral
resistance of the base plate. In fact, the peak loads observed in the load-
deformation plots (Figures 3.23 and 3.24) are an artifact of the loading history
since they coincide with displacements of load reversal. Thus, it is not appropriate
to consider only the observed peak load from the experiments in the development
of predictive models and design approaches. However, the experiments provide
important information regarding overall response of the anchor rod mechanism,
which may be used to inform design considerations. Accordingly, Table 3.7 lists
the peak lateral load observed for each slip direction.
3. For both anchor rod tests, the initial yielding and the post-yield hardening (due to
geometric nonlinear behavior of the rods) are followed by strength as well as
stiffness degradation at repeated cyclic displacement excursions (see Figures 3.23
and 3.24). This degradation may be attributed to damage of the grout in the
31
vicinity the rod under cyclic lateral loading (see Fig. 3.28 for a schematic). The
damage to the grout progressively increases the effective bending length of the
rod, resulting in lower strength and stiffness. In fact, the analysis presented in
Appendix D indicates that this degradation in strength and stiffness can be
simulated with a reasonable degree of accuracy through an assessment of an
appropriate effective bending length.
4. At relatively large lateral base plate displacements, two phenomena occur. First,
increased axial forces are developed in the anchor rods due to large lateral
displacements of the bolted end of the rod. Figure 3.29 illustrates this effect by
plotting the average axial rod strains versus slip for the first 0.8” amplitude
displacement excursion. Second, the inside edge of the base plate hole impinges
on some or all anchor rods (see Fig. 3.28 for a schematic), thereby resulting in a
sudden decrease in the effective bending length of the anchor rod. This produces
the stiffening observed in the specimen response (see Figs. 3.23 and 3.24).
Although the observed stiffening behavior is consistent with tension stiffening (as
would be observed from rod deformation due to large lateral base plate
displacements), the analysis presented in Appendix D indicates that a similar
sudden increase in tension stiffening would become dominant at deformations
well in excess of those observed in the experiments. In addition, the
unsymmetrical nature of the observed response (especially in Test #5; see Figure
3.24), as well as observations from fractured rods (see Figure 3.26) confirms that
the observed stiffening behavior is attributed to the anchor rod bearing within the
base plate hole, rather than within the plate washer (Figure 3.28 illustrates this
effect).
5. Referring back to the load-deformation plots (i.e. Figures 3.23 and 3.24), reversal
of loading produces a pinched hysteretic response as the effective lateral stiffness
of the rods decreases as they approach their original (vertical) position and
increases as they are deformed in the reverse direction. In addition, as discussed
previously, the hysteretic behavior is marked by some strength degradation at
32
repeated lateral displacement amplitudes, likely attributed to damage of the grout

surrounding the rods and/or accumulating plastic strain in the rods themselves.
6. Referring to the load-deformation plots (i.e. Figures 3.23 and 3.24),

unsymmetrical response was observed for both tests. Asymmetry was especially
evident in Test #5 (1-1/4” anchor rod), where the maximum base plate shear load
in one direction was significantly higher (by 79%), as compared to the maximum
shear load in the reverse direction. This unsymmetrical response may be attributed
to the irregular placement of anchor rods within the holes Figures 3.31 and 3.32
illustrate the as-built anchor rod placement for Tests #4 and #5; note the
extremely tight placement of two of the rods for Test #5. When the plate was
pushed in the direction of minimum clearance between the rod and the edge of the
anchor rod hole in the base plate, it made contact with the rod at low lateral plate
displacements. As discussed previously, this caused a reduction in the bending
length of the rod, resulting in stiffening behavior. Figure 3.30 illustrates this effect
schematically. In the reverse direction, the larger clearance between the rod and
edge of the hole delayed this response. In fact, a similar stiffening response is
observed for Test #4 as well – however, due to the placement of the rods (and the
extremely oversized holes), it is more symmetric as compared to the response
observed in Test #5. Recall that the holes in the base plate were sized for the 1-
1/4” diameter anchor rods; the smaller 3/4” rod had a greater hole clearance than
typically prescribed. The effect of the base plate bearing on the anchor rod is
demonstrated by a large kink in the rod at the location of the bottom of the base
plate for Test #5 (see Figure 3.26b). A visual examination of the anchor rods after
testing confirms that the bottom edge of the base plate made contact with at least
two of the small rods (for Test #4) and all four of the large rods (for Test #5).
Appendix C provides the current method for characterizing the anchor rod mechanism
capacity. A more exhaustive analysis of anchor rod response, including an explicit
consideration of second order geometric effects, is provided in Appendix D.
33
Figures 3.33 and 3.34 re-introduce the load-deformation curves for Tests #4 and #5
introduced earlier, with a focus on examining the efficacy of various strength prediction
approaches. As discussed earlier, it is inappropriate (as well as impractical) to accurately
characterize the measured peak loads from the tests. Thus, instead of characterizing the
strength prediction approaches in terms of conventional test-to-predicted strength ratios,
the predicted strengths determined as per various approaches are overlaid on Figs. 3.33
and 3.34 (as well as listed in Table 3.7) for a qualitative assessment. Three methods are
considered to estimate the predicted strengths; two methods consider the interaction of
axial force, bending moment and shear in the anchor rods, whereas the third method
considers the interaction only of axial force and shear. The three strength prediction
methods are now briefly summarized –
1. Method One – strength prediction considering axial force, shear and bending
over a length equal to the thickness of base plate plus half the thickness of the
PVM 1
welded plate washer: This strength estimate Rmeasured , eliptical TPF FPT (See Figs. 3.33 and
3.34) reflects the interaction of axial, shear and flexure stresses in which the
anchor rod is assumed to deform in double (reverse) curvature over a length
corresponding to the distance between the top of the grout pad and the center of
the welded plate washer (see schematic in Figure 3.28). A similar procedure is
suggested by the AISC Design Guide 1 (Fisher & Kloiber, 2006) using the
shear/tension interaction equation (Eq. C-J3-5a) provided in the AISC Steel
Construction Manual (AISC, 2005). A sample calculation using this approach is
provided in Appendix C. In the current analysis, measured (rather than nominal)
PVM
parameters are used for Rmeasured , eliptical . In addition, an elliptical stress interaction
PVM
equation is used. For Test #4, Rmeasured , eliptical is calculated as 10.8 kips, whereas it
is calculated as 35.5 kips for Test #5. From Figures 3.33 and 3.34 these estimates
of strength are lower than the observed strengths. However, two points are
1
TPThe subscript “elliptical” refers to the nature of the material interaction relationship used in the
PT
calculation. The subscript “measured” refers to a consideration of a tabulated rod net tensile area. Later
discussion, including Appendix C, outlines an alternate trilinear material interaction relationship as well as
a consideration of a nominal unthreaded rod area measurement.
34
relevant in this regard. First, as described previously, the high initial strength
observed for Test #4 may be attributed to the presence of grout in the plate holes.
PVM
Once this grout is damaged, Rmeasured , eliptical provides a reasonably conservative
estimate of the strength of the rods. For Test #5, it is especially informative to
consider the negative quadrant of data, which is not affected by the base plate
PVM
bearing on the anchor rod. Even in this case, Rmeasured , eliptical provides a reasonable
and conservative approximation of the strength, not considering the stiffening due
to geometric nonlinearities discussed earlier.
2. Method Two – strength prediction considering axial force, shear and bending
over a length equal to the thickness of grout and base plate plus half the thickness
of the welded plate washer: The second approach is similar to the first one,
except that the assumed effective bending length equals the sum of the thickness
of the base plate, half the thickness of the plate washer, as well as the thickness of
the grout pad. The strength calculated as per this approach (overlaid on Figures
PVM *
3.33 and 3.34) is denoted as Rmeasured , eliptical (calculated as 6.84 kips for Test #4 and
25.0 kips for Test #5 – see Appendix C for a detailed calculation). As expected,
with the increased bending length, the strength estimates are lower (33% on
average) than those predicted by Method One. Consequently, this estimate of
strength is highly conservative with respect to the experimental data and reflects
the observed response only at high levels of base plate slip (approximately greater
than 0.2”) after substantial strength degradation has occurred. This is not entirely
unexpected, given that the strength degradation is associated with grout damage,
and the progressive increase of anchor rod bending length.
3. Method Three – strength prediction considering only axial force and shear: This
PV
estimate of strength Rmeasued , eliptical (which equals 71.1 kips for Test #4 and 158
kips for Test #5) disregards anchor rod bending, but instead considers only the
interaction between axial force and shear (see Appendix C for a detailed
calculation). In addition to these estimated strengths, it is evident from Figures
35
3.33 and 3.34 that Method Three results in the highest estimates of strength. In
PV
fact, Rmeasued , eliptical predicts the peak strength in both anchor rod tests with a higher
degree of accuracy, as compared to Methods One and Two. However, a closer

PV
inspection reveals that Rmeasued , eliptical may provide unreliable (and unconservative)
estimates of shear strength in the base plate connection, since the observed
increase in strength due second order geometric effects requires an assumed
displacement, which may be difficult to quantify for design. Moreover, if
designed for this strength, the anchor rods may undergo large cyclic inelastic
deformation, possibly resulting in fatigue failure.
Based on the above observations, it is evident that bending of the anchor rods should be
considered in the strength determination for shear transfer in a grouted base plate with a
welded plate washer and oversized anchor rod hole detail. A post-test visual inspection of
the anchor rods confirms reversed curvature bending behavior (see Fig. 3.25). Three
other factors may increase the strength of the anchor rod connection, including (1) large
displacements causing the base plate (rather than the welded plate washers) to come into
contact with the anchor rods, resulting in a reduction of the bending length, (2) irregular
placement of the anchor rods, resulting in constrained bending and (3) geometric
nonlinear effects which develops increased tension forces in the deformed anchor rod.
Fortunately, these three effects result in strength capacities which are higher than
capacities based on current design guidelines that consider rod bending. However, the
increased strength associated with these alternate mechanisms depends on factors that
may not be possible to prescribe during design and construction. Thus, based on the
experiments conducted in this study, it is recommended that flexure, as well as axial and
shear stresses, be considered when characterizing the strength of the anchor rods in
bearing. However, incorporating flexural response requires the consideration of an
appropriate bending length. The current approach prescribed in the AISC design guide
(i.e. assuming the bending length equal to the thickness of the base plate plus half the
thickness of the welded washer) provides the most reasonable strength estimate for
design. Although some inelastic deformations are associated with this force level, they
are relatively small.
36
The other alternative bending length considered, (i.e. assuming the bending length equal
to the base plate and grout thickness, plus half the thickness of the plate washer),
provides a somewhat conservative strength prediction. While the strength determined
based on this increased length reflects response at larger displacements (subsequent to
strength degradation), it does not reflect initial response. Thus, this alternative bending
length may be more appropriate under situations of base plate cyclic deformations.
As outlined in Appendix C, all the strength estimates discussed above are based on the
interaction equation presented below between shear and tensile stress –
2 2
⎛ ft ⎞ ⎛ fv ⎞
⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ = 1 [3.1]
⎝ Fnt F
⎠ ⎝ nv ⎠
Where: Fnt = ultimate tensile capacity of rod
f t = tensile stress in rod (from flexure and tensile loads)
Fnv = ultimate shear capacity of rod
f v = shear stress in rod
It is relevant to note, in addition to the interaction equation presented above, the AISC
Steel Construction Manual presents a trilinear interaction equation (Equation [C-J3-6a] in
the Commentary) as an alternative to Equation [3.1]. This trilinear approach is presented
in Appendix C. In addition, the AISC Steel Design Guide 1 assumes the anchor tensile
stress area to be of 75% the nominal (unthreaded) rod area. However, the net tensile area
of the rod may be more accurately characterized by the number of threads per inch and
the measured unthreaded diameter (e.g. see Table 7-18 in the 2005 AISC Manual). The
results presented in Figures 3.33 and 3.34 (and Table 3.7) are based on this more accurate
(i.e. measured) net tensile area of the anchor rod. The use of the alternate (trilinear)
interaction equation and the nominal dimensions will, in general, produce slight
deviations from the predictions presented in this study. While these deviations depend on
37
specific rod size and configuration, for the tests presented in this study, these alternate
capacity predictions produced strength estimates that were similar (within approximately
10%) as compared to the predictions presented in this chapter (i.e. using a measured
diameter and an elliptical stress interaction equation; see Appendix C for details).
The proposed strength prediction method detailed above is significantly conservative as

compared to test observations. The analysis presented in Appendix D indicates that if
geometric nonlinearities are included, strength estimates more consistent with observed
response may be obtained. However, since the force obtained from this analysis is a
function of displacement (being a direct consequence of geometric nonlinearities), it is
difficult to consider it for design, especially for stiff mechanisms, where the deformation
demands cannot be appropriately characterized.
It is relevant to note that, under certain situations, the base plate connection will develop
additional resistance from friction that would develop from clamping action which arises
when the base plate displaces laterally leading to increased tension forces in the rods.
Recall that for the current study, the base plate completely uplifted from the grout pad for
the entire lateral loading protocol. Thus, clamping action of the base plate is beyond the
scope of this report. The reader is referred to Gresnigt et al. (2008) for an analysis that
also includes clamping action of the base plate.

Referring to the test matrix presented in Table 3.4, Tests #6 and #7 feature a concrete
block with a pocket (see Fig. 3.1), into which a shear key (a built-up I-beam stub; see Fig.
3.10) was inserted. These tests investigate the failure modes and capacities associated
with the shear key bearing mechanism. As outlined in the Chapter 2 and in the AISC
Steel Design Guide 1 (Fisher & Kloiber, 2006), shear key failure may be associated with
several failure modes, including concrete/grout bearing failure, shear blowout edge
failure of the concrete foundation or failure of the shear key itself, either through yielding
or through fracture of the welds which connect the shear key to the base plate.
Additionally, the shear key may induce flexural yielding failure in the base plate. In the
38
present study, the main focus is on the shear blowout edge failure of the concrete
foundation. As explained earlier, the predicted failure region of the concrete pedestal
used to represent the foundation did not contain rebar nor any other obstructions (with the
exception of the anchor rods) to enable characterization of the strength capacity
associated with concrete alone (i.e. not including the effect of steel reinforcement). For
both tests, the shear key was installed in the center of the concrete block such that an
edge distance of 20.25” was provided between the shear key bearing surface and the
pedestal perimeter (see Fig. 3.35). Two shear key embedment lengths (measured from
below the surface of the concrete) were tested: 5.5” for Test #6 and 3.0” for Test #7. The
shear key was 6” wide for both tests, thus providing a 33 square inch bearing surface for
Test #6 and 18 square inches for Test #7. A small amount of axial tension force was
applied to the column during both tests to reduce the self weight of the test setup.
However, since the axial loading was applied in load control, some compressive axial
load was maintained during each of the tests (up to 27 kips for all tests; see Table 3.5 and
Appendix B for details). This relatively small compressive load in the column ensured
that the base plate did not completely lift off the concrete pedestal during testing. As a
result, a small amount of resistive frictional force was present, and a corresponding
correction is applied before analysis of the measured shear key strengths.
Under a relatively small axial compressive load, both shear key experiments (Tests #6
and #7) involved the application of lateral (i.e. shear) monotonic loading to failure.
Representative load-deformation plots are shown in Figure 3.36 and key results, such as
observed peak load values, are listed in Table 3.8. Referring to Figure 3.36, load-
deformation plots are shown for the positive and negative loading directions. Data
collected for the reversed loading direction is assumed to reflect undamaged initial
conditions since the concrete behind the shear key’s initial loading direction remained
virtually undamaged. Thus, while the second (i.e. “negative”) load-deformation plot
corresponds to a damaged concrete pedestal, the peak load values (as well as the load-
deformation plot itself) are not substantially different as compared to the load-
deformation plot for the “undamaged” condition of the concrete pedestal (average peak
load difference = 5%).
39
Figure 3.37 shows photographs and schematics illustrating the typical evolution of
damage as lateral loading progressed for both Test #6 and #7. Two distinct load drops are
observed. The first load drop (about 20-40% of the peak load) corresponds to the
formation of a long vertical crack running down the center of the concrete pedestal (see
Figure 3.37a). At this point, cracks were also observed on the grout pad underneath the
base plate. This center crack on side of the concrete pedestal increased in width and
length as the lateral resistance continued to increase (see Figure 3.37b). This failure mode
is similar to flexural response in concrete beams (MacGregor & Wight, 2004), where
flexure cracks perpendicular to the beam axis form before inclined shear cracking leads
to ultimate failure. The absence of steel reinforcement in the failure region precludes
analysis by the strut-and-tie method.
After the initial load drop, the load steadily continued to increase until a second (and
final) peak in load was observed, accompanied by diagonal cracks on both sides of the
pedestal about 30 degrees perpendicular to the loading direction (see Figure 3.37c). This
type of cracking is consistent with shear blowout failure reported in the literature on
concrete design (e.g. MacGregor & Wight, 2004). The post-test condition of the pedestal
of Test #6 is shown in Figure 3.38. Figure 3.39 shows the failure surface of the concrete
pedestal from Test #6 and Test #7 (which was extracted manually after testing).
Appendix B includes a contour plot (generated by three dimensional laser scanning) and
pictures of the concrete failure surfaces for all shear key tests. The steel shear key itself
did not show any signs of damage. In fact, strain data recorded by strain gages on the
shear key (see Fig. 3.11) did not indicate any yielding. The anchor rods, which were
embedded within the failure region, showed permanent deformation in the direction of
loading indicating that they might have contributed to the strength of the failure
mechanism. No concrete or grout bearing damage was observed in the region directly in
front of the shear key’s load path. Note that the final peak load values occur at relatively
small lateral displacements (approximately less than 0.2” for all shear key tests),
indicating the high stiffness of the failure mode associated with this type of detail.
40
Table 3.8 summarizes key experimental results from the shear key tests including the two
initial ultimate
peak lateral loads corresponding to the first ( V peak ) and the second ( V peak ) force drops
in each test. These peak load values are illustrated in the load-deformation plots of Figure
effective
3.36. Referring to Table 3.8, V peak represents the peak load observed during the second
(i.e. ultimate) failure mode minus the friction force. This friction force is determined
from the axial load at the time of the second failure mode (see Table 3.8) using a
coefficient of friction determined in Section 3.4.1 for a steel-grout interface ( μ = 0.45). In
addition to the observed peak load values, Table 3.8 also includes predicted strengths for
the shear key bearing tests. Two methods are considered to estimate the strength
corresponding to the observed blowout failure mode of the concrete. One of these
methods, commonly referred to as the 45 degree cone method, is prescribed by ACI-349
(2006) for the concrete shear capacity of embedded shear lugs and is featured in the
AISC Design Guide 1 (Fisher & Kloiber 2006). As per an adaptation of this method, the
strength capacity corresponding to shear blowout is determined as follows –
( )
Vn45 = 4 f c′ A45 , lb [3.2]
Where: f c′ = concrete compressive strength, psi
A45 = effective stress area based on a 45 degree projected plane, inches2 (seeP P
Figure 3.40a)
The 45 degree cone method assumes a uniform tensile stress of 4 f c′ acting on an
effective stress area defined by projecting a 45 degree plane (note - not a cone surface)
from the bearing edges of the shear lug to the concrete free surface. The bearing area of
the shear lug is excluded from this projected area. Figure 3.40a illustrates the projected
area used for this method, while Appendix E provides detailed calculations determining
this area.
41
The second estimate of strength VnCCD is based on the concrete capacity design (CCD)
method which has been adapted by ACI-349 (2006) for anchorages in concrete under
tension or shear loading. As per this method, the strength is calculated as –
1 ⎛ 40 ⎞
VnCCD = ⎜ f c′ ⎟ A35 , lb [3.3]
c⎝ 9 ⎠
Where: c = free edge distance from the shear key, inches

f c′ = concrete compressive strength, psi
A35 = effective area based on a 35 degree projected plane, inches2 (see Figure
P P
3.40b)
The CCD method assumes an effective area defined by projecting a 35 degree plane from
the bearing edges of the shear lug to the free surface, excluding the bearing area of the
shear lug. A detailed derivation of Equation [3.3], indicating its adaptation from the CCD
method proposed originally by Fuchs et al. (1995), is presented in Appendix E. Referring
to Equations [3.2] and [3.3] above, the main difference between the two strength
equations is associated with the size-effect in concrete derived from fracture mechanics
theory (Bažant, 1984). Equation [3.2] predicts that the strength associated with concrete
blowout is directly proportional to the projected effective stress area. While this is often
true for smaller specimens, where the specimen dimension is approximately 10-20 times
the aggregate size (Bažant, 1984), failure in larger specimens is typically governed by
fracture mechanics, since the initiation of cracking in the concrete, rather than the
development of a uniform stress over a failure surface, controls the strength associated
with blowout. Similar analogies may be found in other structural components, for
example in steel, wherein some situations and details with a sharp crack do not achieve
their full strength as would be predicted through an ultimate strength approach that
considers only yielding. Research by Bažant (1984) and others (Fuchs et al., 1995) has
shown that this size effect may be successfully incorporated by expressing the “nominal”
failure stress as a function of the specimen size (conveniently characterized by the
42
embedment free edge distance c ), i.e. σ failure = f (1 c, f c′ ) . Thus, for larger
specimens, the failure stress is smaller as compared to that for relatively smaller
specimens. As per the CCD method, the projected area for determination of the nominal
stress is based on a 35 degree projection as illustrated in Figure 3.40b. The reader is
referred to Fuchs et al. (1995) for more details regarding the CCD method.
Along with the experimental data outlined earlier, Table 3.8 includes predicted strength
capacities and test-to-predicted load ratios based on the two estimates described above. In
the context of the above discussion, a review of Table 3.8 and Figs. 3.36-3.39 reveals
several points that are relevant to the strength characterization due to shear key bearing
and concrete blowout failure –
1. Referring to the representative shear load versus base plate slip response (Figure
3.36) and the failure progression photographs (Figure 3.37), the two load peaks are
associated with distinct failure modes; (1) flexural splitting of the free edge followed
by (2) side blowout. For the positive loading direction (i.e. loading on the undamaged
pedestal), the second failure mode corresponds to the largest capacity for both tests.
Although the initial flexural splitting failure mode corresponds to a drop in load,
ultimate failure corresponds to concrete blowout in a shear mode. Thus, the loads
effective
corresponding to this event, corrected for friction (i.e. V peak ), are included in the
analysis of the results and in further discussion.
effective
2. An inspection of the test-to-predicted ratios V peak / Vn45 (mean = 0.51; COV = 0.14)
indicates that the 45 degree cone method (currently proposed by the AISC Design
Guide 1) may be unconservative for large edge distances (i.e. large concrete
foundations) due to the size effect discussed in the preceding discussion.
3. The test-to-predicted ratios determined with respect to the CCD method indicate the
inclusion of the concrete size effect in the analysis provides more accurate estimates
effective
of strength (mean V peak / VnCCD = 1.07; COV = 0.19).
43
4. A slight bias is observed such that both the 45 degree cone method as well as the
CCD method gives higher strength predictions for the deeper 5.5” shear key (mean
effective
V peak effective
/ VnCCD = 1.22; mean V peak / Vn45 = 0.55) as compared to the 3.0” deep shear
effective
key (mean V peak effective
/ VnCCD = 0.92; mean V peak / Vn45 = 0.46). Due to the geometry of
the concrete pedestal, the effective stress area A35 is nearly identical for the two shear
key lengths (less than 1% difference), while the area A45 differs by about 10% for the
two shear key embedment lengths. This suggests that the longer shear key is stronger
on a unit basis, as compared to the shorter one. This may be attributed to the larger
local bearing stresses associated with the smaller shear key embedment length. As
evidenced in numerical studies (e.g. Ožbolt et al., 2007), this higher bearing stress
increases the likelihood of early crack initiation and failure.
5. For both tests, the angle of the failure surface in plan view (i.e. the top of the
pedestal) is approximately 30 degrees perpendicular to the load (e.g. see Figs. 3.37
and 3.38). In elevation view (i.e. the side of the pedestal), the angle of the failure
surface is between 40-60 degrees to the load (e.g. see Figure 3.39). Thus, the overall
failure surface area is within the range predicted by the 45 degree cone method and
the CCD method.
Based on the above observations, the CCD method provides a relatively accurate estimate
of strength for embedments with large free edge clear distances (i.e. large concrete
foundations), where the strength is controlled by fracture initiation. Although not tested
as part of this study, it is anticipated that for small edge distances, where the strength is
governed by development of the concrete tensile strength over the failure area, the CCD
method might provide unconservative results. Thus, it is recommended that the reliable
strength of concrete blowout due to shear key bearing be calculated as the minimum of
the two estimates, such that for smaller edge distances, the 45 degree cone method will
govern, while for larger edge distances, the CCD method will govern –
44
⎧
φ Vn = min⎨φCCD
1 ⎛ 40
⎜
c⎝ 9
⎞
( ) ⎫
f c′ ⎟ A35 , φ45 4 f c′ A45 ⎬ [3.4]
⎩ ⎠ ⎭
Where: φ CCD = strength reduction factor of the CCD method = 0.75 (ACI 349, 2006)
φ 45 = strength reduction factor of the 45 degree cone method = 0.85 (ACI 349,
2006)
The two functions in Equation [3.4], along with the observed effective peak loads, are
plotted in Figure 3.41 for clarity. Referring to this equation, two important observations
are presented. First, the effective stress areas calculated as per the two methods are based
on different geometries, i.e. one is based on a 45 degree angle, whereas the other (i.e. the
CCD method) is based on a 35 degree angle (see Figure 3.40). The area calculated for the
45 degree cone method is based on physical interpretation, since this method assumes
that the concrete tensile strength is activated over this area. On the other hand, the
expression for the CCD method is based on fracture mechanics, i.e. all of the projected
area is not simultaneously engaged. In fact, in situations where the CCD method governs,
failure occurs due to the initiation of cracking over a small region in the vicinity of the
shear key. Once crack initiation occurs, the load drops steadily as the shear cracks grow
and the failure region expands. Thus, the area A35 in Equation [3.3] above does not bear
any physical significance in the context of the final failure surface, but rather may be
interpreted as a basis for the characterization of the nominal stress required to produce
fracture. As for the second observation, assuming that the projected areas can be
calculated based only on the edge distance c (i.e. the dimension of the shear key is small
as compared to the concrete foundation and the foundation is large enough to capture the
entire projected area), Equation [3.4] above indicates that the CCD method will govern
for edge distances greater than approximately 6 inches (i.e. 5-10 times the aggregate size,
which is in approximate agreement with the fracture mechanics theory proposed by
Bažant [1984]).
The size effect in concrete is diminished in the presence of steel reinforcement which
increases the ductility of the concrete pedestal, thereby providing the opportunity for the
45
redistribution of stresses over a larger volume of material. However, the beneficial effect
of reinforcement is difficult to quantify in the absence of additional testing.
In some cases, the shear key may be positioned close to a corner, or the foundation may
not be deep enough to develop the full projected effective area (i.e. A45 or A35 ).
Appendix E outlines methods and provides schematics for calculating these areas in the
presence of these boundary effects.
46
Table 3.1 – Summary of anchor rod tension tests

Nominal unthreaded Fy,rod1
B PB P Fu,rod2
B PB P E3 P P
diameter (inches) (ksi) (ksi) (ksi)

3/4 66.8 96.4 32,000
1-1/4 54.4 75.0 31,100
1
P Measured yield stress, based on 0.2% offset method
P
2
P Measured ultimate strength
P
3
P Measured Young’s modulus
P
Table 3.2 – Summary of concrete compression tests

Concrete Corresponding Concrete Average 28-day Standard deviation
Delivery
pedestal base plate test density compressive of compressive
truck
specimen1 P P number (lbs/ft3)
P P strength (psi) strength (psi)
A 1,2,3,4 A 146 4,570 290
B 5 A 146 4,760 60
C 6 A 145 4,650 320
D 7 B 145 5,030 90
1
P Three 6”-by-12” test cylinders were collected for each concrete pedestal specimens
P
Table 3.3 – Summary of grout compression tests

Corresponding Water content Curing Grout Average Standard deviation
base plate test (gallons per time density compressive of compressive
number1 50 pounds of grout)
P P (days) (lbs/ft3) P P strength (psi) strength (psi)
1 1 9 128 5,780 1,050
2 1 13 127 7,070 760
3 1 7 128 6,230 390
4 7/8 4.5 130 6,130 1,100
5 7/8 7 131 7,210 970
6 1-1/8 9 126 6,030 850
7 1-1/8 7 127 5,800 1,050
1
P Four 3”-by-6” test cylinders were collected from the grout for each base plate test
P
Table 3.4 – Base plate test matrix

Mechanism
Test # Test detail Loading description
investigated
1 Cyclic shear with various
Surface Shim stacks plus grout
2 levels of constant axial
friction
3 Grout only compression
3/4” diameter anchor rods with
4
Anchor rod welded plate washers Cyclic shear with constant
bearing 1-1/4” diameter anchor rods axial tension
5
with welded plate washers
6” bearing width with 5.5”
6
Shear key embedment depth Monotonic shear with small
bearing 6” bearing width with 3.0” compressive axial load
7
embedment depth
47
Table 3.5 – Base plate test loading details

Test Loading Axial load Axial loading
Lateral (shear) loading
# iteration (kips) type
1 43
1 2 112
3 261
1 261
Constant Cyclic displacement
2 2 112
compression (see Figure 3.13)
3 43
1 261
3 2 112
3 43
4 1 40 Constant Cyclic displacement to failure
5 1 108 tension (see Figure 3.13)
1 0 to 10 Monotonic push to failure
6 Monotonic push in opposite
2 17 to 27
Small direction to failure
1 15 compression Monotonic push to failure
7 Monotonic push in opposite
2 15
direction to failure
Table 3.6 – Base plate surface friction test results

Grout Corresponding Axial compression Shear force Coefficient of
detail test number (kips) (kips) friction
1 43.0 17.51 0.41
With shim
2 112 44.51 0.40
stacks
2 261 125 0.48
3 43.0 16.5 0.38
No shim
3 112 50.0 0.45
stacks
3 261 119 0.46
48
Table 3.7 – Anchor rod bearing test results

Test Number 4 5
Nominal unthreaded rod diameter (inches) 0.75 1.25
Measured threaded rod diameter (inches) 0.64 1.10
Measured rod ultimate strength (ksi) 96.4 75.0
Plate washer thickness (inches) 0.25 0.50
Base plate thickness (inches) 2.00 2.00
Average grout height (inches) 1.25 1.00
Imposed tensile axial load (kips) 39.6 108
R peak (kips)1 B B
30.2 126
reverse
R peak (kips)2 B B
28.2 70.4
PVM
R measured , eliptical (kips)3 B B
10.8 35.5
PVM *
R measured , eliptical (kips)4 B B
6.84 25.0
PV
R measued , eliptical (kips)5 B B
71.1 158
1
P Observed peak load
P
2
P Observed peak load in opposite slip direction
P
3
P Predicted capacity assuming rod tension, shear and flexure and a bending length equal to the base plate
P
thickness plus half the plate washer thickness

4
P Predicted capacity assuming rod tension, shear and flexure and a bending length equal to the base plate
P
thickness, grout thickness plus half the plate washer thickness P
5
P Predicted capacity assuming rod tension and shear
P
49
Table 3.8 – Shear key bearing test results

Test Number 6 7
1
Shear key embedment
5.5 3.0
P P
length (inches)
Loading direction positive negative positive negative
2
P Axial load (kips)
P 2.0 28.5 15.4 15.4
Frictional resistance (kips) 0.9 12.8 6.9 6.9
3 initial
P P
V peak (kips) 186 185 141 135
4 ultimate
P P
V peak (kips) 194 172 149 143
5 effective
P P
V peak (kips) 193 159 142 136
6
P P
Vn45 (kips) 318 318 302 302
effective
V peak Vn45 0.61 0.50 0.47 0.45
7
P P
VnCCD (kips) 144 144 151 151
effective
V peak VnCCD 1.34 1.10 0.94 0.90
1
P Shear key width = 6”; edge distance = 20.25”
P
2
Axial load at final (ultimate) peak lateral load
3
Observed initial peak load
4
Observed second (ultimate) peak load
5
Observed ultimate peak load adjusted for frictional resistance
6
Predicted capacity using 45 degree cone method
7
Predicted capacity CCD method
50
Figure 3.1 – Concrete pedestal with anchor rods and shear

key pocket
(a) (b)
Beam-Column
Out-of-Plane
Bracing
bearing
23’
Actuators
Base
Plate
Concrete
Pedestal
Figure 3.2 – Base plate shear test setup (a) schematic (to scale; rear reaction wall
omitted for clarity) and (b) photograph
51
Steel Assembly
for Shear
Application
Column
Base
Plate
Actuator
Concrete Pedestal
Figure 3.3 – Steel shear loading assembly for base plate tests
Figure 3.4 – Steel shim stacks used for friction Tests #1 and #2
52
Foam
(a)
Dam
(b) Shim
Stack
Loading
Direction
Figure 3.5 – Shim stack positioning for (a) Test #1 and (b) Test #2
Loading Direction
Figure 3.6 – Post-test photograph of grout

pad for Test #3
53
(a)
(b)
Figure 3.7 – Schematic (a) plan view and (b) elevation view of reinforcing bar details
used in the concrete pedestals for the anchor rod tests #4 and #5
54
Connection for Actuator W31x191 Column Connection for Lab Floor
Anchor Rod Hole Extent of Grout

for Friction Tests
Extent of Grout
for Anchor Rod
and Shear Key
Tests
24”
Concrete Pedestal 24” Base Plate
Figure 3.8 – Schematic plan view of base plate specimen (to scale)
Figure 3.9 – Welded plate washer detail (shown here

for Test #5)
55
Loa d
in g D ir
ec tio n
Base Plate
Figure 3.10 – Welded shear key (shown here for Test #6)
7.5”
6”
3” 5.5”
2.5” Test #6: 7”

2.5” 5” Test #7: 4.5”
5”
Strain
Gage
Strain
Gage
Figure 3.11 – Schematic detail of shear key geometry
56
(a)
(b)
Figure 3.12 – Schematic (a) plan view and (b) elevation view illustrating
rebar details in the concrete pedestals for the shear key tests (Test #6 and #7)
57
1.0
0.8
0.6
Lateral Displacement, inches
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
-1.0
Time
Figure 3.13 – Typical lateral loading protocol for large scale surface friction and anchor
rod bearing tests (Tests #1 - #5)
58
(a)
UP
Legend
U
E W
In-plane
displacement gage
DOWN Out-of-plane
displacement gage
Strain gage
(b) S
E W
Figure 3.14 – Schematic illustrating instrument locations (to-scale)
59
300
Steel-grout bond rupture

Frictional Force, kips
-300
-1.1 0.0 1.1
Figure 3.15 – Representative response plot of the surface friction tests

(shown here for Test #3)
60
(a)
(b)
(c)
Figure 3.16 – Condition of (a) the underside of the base

plate, (b) the grout pad and (c) shim stacks after Test #1
61
300
axial load = 261 kips axial load = 112 kips axial load = 43 kips
150
0
0 30 60 90
Cumulative Slip, inches
Figure 3.17 – Frictional force versus cumulative slip for Test #2
62
(a) Test #1
1.0
Effective Coefficient of Friction

axial load = 43 kips axial load = 112 kips axial load = 261
0.5
μmin = 0.40
μmin = 0.41
0.0
0 30 60 90
(b) Test #2
1.0
0.5
μmin = 0.48
0.0
0 30 60 90
(c) Test #3
1.0
0.5
μmin = 0.45
μmin = 0.46 μmin = 0.38
0.0
0 30 60 90
Figure 3.18 – Response progression of each friction test illustrating the

extracted coefficient of friction values used for analysis
63
150
With Shim Stacks
No Shim Stacks
46 (With Shim Stacks)
μ=0.45
μ=0.45 (No Shim Stacks)
75
0
0 150 300
Axial Compression, kips
Figure 3.19 – Summary of friction tests illustrating the extracted data

points used to determine the coefficient of friction
64
120
Test #4
Test #5
80
Axial Load, kips
40
-40
0.00 0.04 0.08
Vertical Uplift, inches
Figure 3.20 – Vertical uplift versus axial load during the initial
axial tension loading of the anchor rod bearing Tests #4 and #5
65
50
NE Rod SE Rod NW Rod
SW Rod
Total Axial Load (kips)
-50
0 500 1,000
Rod Axial Strain (με)
Figure 3.21 – Anchor rod axial strains during initial axial uplift of base plate Test #4
(3/4” diameter rods)
66
0.6
Test #4
Test #5
0.3
0.0
0 10 20
Cumulative Lateral Slip, inches
Figure 3.22 – Vertical displacement of base plate versus cumulative lateral
slip during lateral loading of both anchor rod bearing Tests #4 and #5
67
40
reverse
Initial Peak Load
Stiffening Due R peak
to Base Plate
from Grout
Contact
Interference
Lateral Load, kips
Rod Fracture
Strength Degradation
Due to Grout Damage
R peak
-40
-1 0 1
Figure 3.23 – Load versus displacement response of Test #4 – 3/4” diameter anchor rod
140
Strength Degradation R peak
Due to Grout Damage
Lateral Load, kips
Stiffening Due to Rod Fracture

R reverse Base Plate Contact
peak
-140
-1 0 1
Figure 3.24 – Load versus displacement response of Test #5 – 1-1/4” diameter anchor rod
68
(a) (b) (c)
Figure 3.25 – Post-test photographs showing (a) rod fracture of Test #4, (b) fractured
3/4” diameter rod from Test #4 and (c) fractured 1-1/4” diameter rod from Test #5
69
Unwelded Welded
(a) Washer Washer
Two
Base Plate
Inches
Grout
Fracture Concrete
Location
(b)
Unwelded
Washer Welded Washer
Two
Inches Base Plate
Grout
Fracture Concrete
Location
Figure 3.26 – To-scale illustration of fractured (a) 3/4” diameter

anchor rod and (b) 1-1/4” rod relative to base plate, grout and
concrete
70
(a)
(b) (c)
Figure 3.27 – Post-test photographs showing (a) grout damage of Test #4; (b)
concrete damage of Test #5 and (c) grout damage of Test #5
71
Undeformed Rod
U
anchor rod
nut
welded
washer
base plate
grout
concrete
Plate Washer Bearing

U Damage to Grout
U
L = half plate L = half plate

washer thickness washer thickness
plus base plate bearing location plus base plate and
thickness grout thickness
load load
damage to grout
Tension Stiffening
U Plate Bearing
U
load load
tensile force in rod L = grout thickness bearing location
Figure 3.28 – Schematic illustrating various phenomenon in the anchor rod

bearing mechanism
72
400
Test #5
Average Rod Strain, μstrain
Test #4
Average Rod Strain
200
0
0.0 0.5 1.0
Figure 3.29 – Average (relative) axial strain measured from all anchor rods during the
first lateral displacement excursion to 0.8”
73
anchor rod
welded washer
base plate
grout
concrete
Higher Force
Leffective
B
bearing location
bearing location
Leffective
B
Lower Force
Figure 3.30 – Schematic illustrating the mechanism causing the

asymmetry in anchor rod bearing load-deformation response
74
Anchor Rod
Base Plate
(a) Column
Loading
Direction Fractured Rod
Anchor Rod
Column
(b)
Fractured Rod
Figure 3.31 – (a) Superimposed photographs and (b) to-scale schematic indicating as-
built anchor rod positioning for Test #4 (3/4” diameter rods)
75
Base Plate
Anchor Rod
(a) Column
Loading
Direction Fractured Rod
Anchor Rod
Column
(b)
Fractured Rod
Figure 3.32 – (a) Superimposed photographs and (b) to-scale schematic indicating as-
built anchor rod positioning for Test #5 (1-1/4” diameter rods)
76
40
Experimental Response
PVM
Prediction Rmeasued , eliptical
PVM *
Prediction Rmeasued , eliptical
Lateral Load, kips
-40
-1 0 1
Figure 3.33 – Anchor rod strength estimates and observed response for Test #4 (3/4”
diameter anchor rod); PV estimate out of range
140
Experimental Response
Prediction Rmeasued
PVM
, eliptical
PVM *
Prediction Rmeasued
Lateral Load, kips
, eliptical
-140
-1 0 1
Figure 3.34 – Anchor rod strength estimates and observed response for Test #5 (1-1/4”
diameter anchor rod); PV estimate out of range
77
39”
6”
39” 20.25”
7.5”
20.25”
5.5”
or 3.0”
32.5”
Shear Key
48” 84”
Figure 3.35 – Dimensions of concrete pedestal with shear key pocket
78
200 initial
V peak
ultimate
V peak
Lateral Load, kips
100
(a)
0
0.0 0.2 0.4
Lateral Slip, inches
200
initial ultimate
V peak V peak
Lateral Load, kips
100
(b)
0
0.0 0.2 0.4
Lateral Slip, inches
Figure 3.36 – Representative load-displacement response for the shear

key tests; shown here for Test #6 – 5.5” shear key depth (a) the positive
loading direction and (b) the negative loading direction
79
(a)
shear load
shear load
(b)
(c)
shear load
≈ 30º
Figure 3.37 – Damage progression of initial loading of Test #6 at (a) 0.066” slip
(b) 0.144” slip and (c) 0.246” slip (refer Figure 3.36a)
80
(a)
(b)
Figure 3.38 – Post test photographs of Test #6 (a) elevation view (b) plan view
81
(a)
(b)
Figure 3.39 – Failure surface of (a) Test #6 and (b) Test #7
82
(a)
shear load
c
w
45º
d+c
A45 B B
w + 2c
(b) shear load

c
w
35º
d + 1.5c
A35
B B
w + 3c
Figure 3.40 – Schematic illustrating the effective stress area assumed by (a)
the 45 degree cone method and (b) the CCD method (c = edge distance; w =
shear key width; d = shear key embedment depth)
83
300
45 Degree Cone Method
CCD Method
Shear Capacity, Vn (kips)
Test #6 Effective Peak Load

Test #7 Effective Peak Load
150
0
0 10 20 30
Free Edge Distance, c (inches)
Figure 3.41 – Plot relating the shear capacity of a shear key embedded in concrete
( f c′ = 4000 psi) a distance “c” to a free edge using the 45 degree cone method and
the CCD method; experimental observations included
84
Chapter 4: Summary and Conclusions
Chapter 4
Summary and Conclusions
4.1 SUMMARY
Current methods and equations for the design of shear transfer in steel column bases,
such as those presented in the AISC Steel Design Guide 1 (Fisher & Kloiber, 2006), have
not been fully validated by experimental testing. An independent review of literature,
including review of a recently published AISC sponsored synthesis of research on
column base connections (Grauvilardell et al., 2005), reveals that large scale
experimental tests of shear transfer in exposed column base plates has not been explicitly
addressed, although associated mechanisms have been researched in independent
contexts. Thus, current design provisions for base plates typically adapt and combine
findings from several independent studies, which are, in general, based on small-scale
component tests or specific failure mechanisms. Therefore, only limited data is available
involving large scale base plate components, where various mechanisms may interact
with each other or may be influenced by the construction procedure or geometry of the
base connection itself.
To address these issues, an experimental study was conducted to investigate three

common shear transfer mechanisms of exposed column base plates which are featured in
the AISC Steel Design Guide 1 including (1) surface friction (2) anchor rod bearing and
(3) shear key bearing. Seven large scale base plate tests were carried out; three tests
investigated surface friction between the base plate and grout interface (two tests
integrated steel shim stacks while one test examined steel on grout), two tests
investigated anchor rod bearing (3/4” and 1-1/4” diameter rods) and two tests
investigated shear key bearing (5.5” and 3.0” shear key embedment depth below the
surface of the concrete). Loading conditions included axial compression, axial tension
and monotonic/cyclic shear. The base plate tests were supplemented by ancillary tests to
measure anchor rod, concrete and grout properties.
85

The three base plate tests investigating surface friction involved the application of three
different compressive axial loads levels (43, 112, and 261 kips; with corresponding
bearing stresses of about 64, 166, and 386 psi) between the base plate and grout interface.
Two tests included shim stacks installed in the grout pad while a third test did not contain
shims. For each constant compressive axial load level, a lateral displacement-based load
protocol was applied. The general behavior for all tests included a significantly large
initial lateral resistance due to the steel-grout bond (from chemical adhesion) followed by
a square shaped hysteresis plot indicative of Coulomb friction behavior. For the tests
featuring the steel shim stacks, frictional resistance ultimately increased as lateral loading
cycles continued, most likely attributed to the shims gouging into the base plate. The
shear force corresponding to initial bond breakage and shim gouging are ignored for
evaluation of the friction coefficient since these mechanisms may not be present in actual
base plate conditions and may provide unconservative strength estimates.
Based on the experimental data, a coefficient of friction value of 0.45 is recommended

for use in design. This value is lower than previously reported values and 20% lower than
the design value featured in the AISC Design Guide One (0.55). Friction, a complex
phenomenon in general, is sensitive to several factors, such as the properties of the grout
and the surface condition of the steel base plate. The presence of mill scale, such as
included in this current study, may result in lower coefficient of friction values compared
to a machined steel surface.

Two base plate tests investigated the shear resistance of four anchor rods under a
combination of imposed axial tensile loads and shear/flexural loading. The connection
included welded plate washers and grout between the base plate and concrete pedestal.
Each test examined a different nominal anchor rod diameter size; i.e. 3/4” and 1-1/4”.
The lateral load versus lateral plate displacement response is similar for both tests. The
cyclic loading resulted in immediate nonlinear response of the connection including
86
stiffening in the rods resulting from geometric nonlinearities introduced due to large
lateral displacements of the bolted end of the anchor rod. A pinched hysteretic behavior is
observed as well as some strength degradation at repeated lateral displacement
amplitudes. Significant grout damage, in the form of cracking and spalling around the
anchor rods was observed for both tests. Both tests were concluded when one rod
fractured during a cyclic displacement excursion. Asymmetrical response was observed
for both tests, attributed to the irregular placement of anchor rods in the holes during
construction which induced constrained rod bending in one loading direction.
Three methods were evaluated to calculate the anchor rod bearing strength. Two of these
methods incorporate the interaction of axial, shear and flexure stresses while another
method considers only the interaction between axial stress and shear. Based on
experimental observations, the method that neglects the effect of flexure is determined to
be significantly unconservative. Thus, it is recommended that flexure be considered in the
design of anchor rods to resist shear.
The main difference between the two strength prediction methods that consider flexure is
the effective bending length, over which the anchor rods are assumed to bend in reverse
curvature (based on visual inspection of the fractured rods). One of these methods (based
on the current approach prescribed in the AISC Steel Design Guide 1) assumes that the
bending length equals the thickness of the base plate plus half the thickness of the welded
plate washer. Analysis of experimental evidence indicates that this assumption is
consistent with observed response at small base plate slip levels. However, as base plate
slip increases, damage to the grout surrounding the anchor rod results in an increase in
bending length, leading to strength degradation during subsequent cyclic loading. A third
strength prediction method (which also considers flexure) assumes that the rods bend
over a larger length: the thickness of the base plate and the grout height plus half the
thickness of the welded plate washer. This assumption provides a somewhat conservative
strength prediction. While the strength determined based on this longer length reflects
response at larger lateral base plate displacements (subsequent to strength degradation), it
does not reflect initial response of the connection.
87
At larger displacements, tension stiffening due to deflection of the bolted end of the
anchor rods (as well as rod contact within the base plate holes) result in a significant
increase in strength. Considering these effects, the current approach prescribed in the
AISC Design Guide (i.e. assuming the bending length equal to the thickness of the base
plate plus half the thickness of the welded washer) is determined to provide a reasonably
conservative strength estimate for design.

Two base plate tests featured a pocket within a concrete pedestal, into which a shear key
was inserted, to investigate the failure modes and capacities associated with a shear key
bearing mechanism. Two shear key embedment lengths were investigated (5.5” and 3.0”
below the surface of the concrete). The present study focused on the shear blowout edge
failure mechanism of the concrete foundation. The predicted failure region of the
concrete pedestal used to represent the foundation did not contain rebar nor any other
obstructions (with the exception of the anchor rods) to enable characterization of the
strength capacity associated with concrete alone (i.e. not including the effect of steel
reinforcement). Under small axial compressive loads, both shear key experiments
involved the application of lateral (i.e. shear) monotonic loading to failure. Two distinct
force peaks are observed. The first force peak, accompanied by a sudden drop in load
(between 20-40% of the peak load), corresponds to flexural splitting of the concrete
pedestal. After the initial drop in load, the load steadily increased until a second (and
final) peak in load was observed. This ultimate peak in load was accompanied by
diagonal cracks on both sides of the pedestal, about 30 degrees perpendicular to the
loading direction.
Two methods are considered to estimate the strength corresponding to the observed
blowout failure mode of the concrete. One of these methods, commonly referred to as the
45 degree cone method, is prescribed by ACI-349 (2006) for the concrete shear capacity
of embedded shear lugs and is featured in the AISC Design Guide 1 (Fisher & Kloiber,
2006). A second estimate of strength is based on the concrete capacity design (CCD)
88
method, which has been adapted by ACI-349 (2006) for anchorages in concrete under
tension or shear loading. The main difference between the two strength prediction
methods is associated with the size-effect in concrete derived from fracture mechanics
theory (Bažant, 1984). An inspection of the test-to-predicted ratios indicates that the 45
degree cone method (currently proposed by the AISC Design Guide 1) may be
unconservative for large concrete edge distances (i.e. large concrete foundations) due to
this size effect. The 45 degree cone method has a mean test-to-predicted strength ratio of
0.51 (COV = 0.14). The test-to-predicted ratio determined with respect to the CCD
method indicates the inclusion of the size effect in the analysis provides a more accurate
estimate of strength. In this case, the mean test-to-predicted ratio is equal to 1.07 with a
COV equal to 0.19. The longer shear key is observed to be stronger on a unit basis, as
compared to the shorter shear key. The larger local bearing stresses of a shorter lug may
increase the likelihood of early crack initiation.
Based on experimental observations, the CCD method provides a relatively accurate

estimate of strength for embedments with large free edge clear distances (i.e. large
concrete foundations), where the strength is controlled by fracture initiation. Although
not tested as part of this study, it is anticipated that for small edge distances, where the
strength is governed by development of the concrete tensile strength over the failure area,
the CCD method might provide unconservative results. Thus, it is recommended that the
reliable strength of concrete blowout due to shear key bearing be calculated as the
minimum of the two previously described estimates.
4.2 CONCLUSIONS AND DESIGN IMPLICATIONS

Based on the reported test data and consideration of previously published literature, this
study has resulted in improved (and more reliable) guidelines and equations to
characterize the strength capacity for various failure modes associated with shear transfer
mechanisms in exposed column base plates. Examples illustrating the use of these
improved methods are included in Appendices C and E.
While these findings may be suitably incorporated into improved design guidelines for
shear transfer, it is outside the scope of this preliminary report, whose main objective is
89
to present the experimental data and associated analyses. Future efforts, including
collaboration with the AISC oversight committee, could result in specific design
guidelines (in the form of reports or papers) or modifications to the AISC Steel Design
Guide 1.
The incorporation of these findings into design guidelines may involve a consideration of
several issues, some of which are outlined below –
1. It is important to note that owing to the expense associated with large-scale testing,
the current data set does not include replicate data sets for statistical analysis. Thus, to
apply this data in a design context, appropriate resistance factors ( φ ) should be
developed through examination of previous standards, specifications and similar test
data.
2. Some of the tests and analyses conducted in this study did not lead to definitive
methods to characterize the strength capacity. For example, from the anchor rod tests,
the geometric nonlinear effects associated with large displacements, combined with
the variation in bearing conditions (including those affected by construction
tolerances), resulted in significantly higher strength capacities than predicted by a
consideration of axial, bending and shear stresses in the rods. However, it is
challenging to characterize this strength since it depends on the deformation capacity
of the anchor rods, which may be difficult to characterize accurately. It may be
desirable to utilize this overstrength for design, if suitable assumptions can be made
regarding the deformation capacity of the anchor rods.
3. In field details, one or more of the mechanisms discussed in the study could be active.
Initially, the force will be resisted by the “stiff” mechanisms of friction (if
compressive load is present) and shear key bearing (if a shear lug is provided). Once
these are overcome, the lateral load may be resisted by other more “flexible”
mechanisms, such as anchor rod bearing. Based on the findings of this and other
90
studies, design provisions may address the combination of various shear transfer
mechanisms to resist the applied loads.
In addition to the three mechanisms outlined in this study, several other details are
currently used in design practice to resist large shear forces in base plates, such as
commonly encountered in high-seismic regions. These alternative details may include
embedding the base plate in concrete or attaching the base plate to a grade beam. Design
guidelines may include a consideration of these details and appropriate methods to use
them in combination with the mechanisms investigated in this study.
Finally, as discussed in the introduction, the experimental results presented in this report
are one part of a comprehensive study on column base plates. While this phase of testing
focused on shear dominated response, in general, behavior will involve both flexure and
shear. Ongoing work focuses on flexural response of base plates, through a series of
seven large scale experiments and complementary simulations. Various parameters
investigated in the ongoing study include base plate thickness, anchor rod layout and
gravity load effects. Although these experiments focus on flexure-dominated response, in
the context of this report, they may provide additional information regarding the
interactive effects of flexural and shear response.
91
References
References
ACI Committee 209. (2006), “Prediction of Creep, Shrinkage and Temperature Effects in
Concrete Structures,” Farmington Hills, MI.
ACI Committee 318. (2002), “Building Code Requirements for Structural Concrete (ACI
318-02) and Commentary (ACI 318R-02),” Farmington Hills, MI.
ACI Committee 349. (1985), “Code Requirements for Nuclear Safety Related Concrete
Structures and Commentary (ACI 349-85/349R-85),” Farmington Hills, MI.
Adihardjo, R., and Soltis, L. (1979), “Combined Shear and Tension on Grouted Base
Details,” Engineering Journal, AISC, Vol. 16, No. 1 (First Quarter), 1979, pp. 23-
26.
American Institute of Steel Construction (AISC). (2006), “Fundamentals of Column

Bases and Exposed Seismic Base Design,” Request for Proposal (RFP) 6807,
August 11, 2006.
American Institute of Steel Construction (AISC). (2005), “Seismic Design Manual,”

American Institute of Steel Construction, Chicago, IL.
American Institute of Steel Construction (AISC). (2005), “Steel Construction Manual,”

13th Edition, American Institute of Steel Construction, Chicago, IL.
92
References
Astaneh-Asl, A. (2008), “Seismic Behavior and Design of Base Plates in Braced

Frames,” SteelTIPS, Technical Information and Product Service, Structural Steel
Educational Council.
Astaneh-Asl, A., Bergsma, G. (1993), “Cyclic Behavior and Seismic Design of Steel
Base Plates,” Proceedings, Structures Congress, ASCE 1993; 409–14.
ASTM Standard C 31. (2008), “Standard Practice for Making and Curing Concrete Test
Specimens in the Field,” ASTM International, West Conshohocken, PA.
ASTM Standard C 39. (2005), “Standard Test Method for Compressive Strength of
Cylindrical Concrete Specimens,” ASTM International, West Conshohocken, PA.
ASTM Standard C 109. (2007), “Standard Test Method for Compressive Strength of
Hydraulic Cement Mortars (Using 2-in. or [50-mm] Cube Specimens),” ASTM
International, West Conshohocken, PA.
Bailey, J.W., and Burdette, E.G. (1977), “Edge Effects on Anchorage to Concrete,”
Research Series No. 31, Aug. 1977, Department of Civil Engineering, University
of Tennessee, Knoxville.
Ballio, G., and Mazzolani, F.M. (1983), “Theory and Design of Steel Structures,”
Chapman and Hall, London and New York, pp. 257-264.
Baltay, P., and Gjelsvik, A. (1990), “Coefficient of Friction for Steel on Concrete at High
Normal Stress,” Journal of Materials in Civil Engineering, Vol. 2, No. 1,
February, pp. 46-49.
Bažant, Z.P. (1984), “Size Effect in Blunt Fracture: Concrete, Rock, Metal,” Journal of
Engineering Mechanics, ASCE, 110(4), 518–535.
93
References
Burda, J.J., and Itani, A.M. (1999), “Studies of Seismic Behavior of Steel Base Plates,”
Report No. CCEER 99-7, Reno (NV): Center of Civil Engineering Earthquake
Research, Department of Civil and Environmental Engineering, University of
Nevada, NV.
Burdette, E.G., Perry T.C., and Funk, R.R. (1987) “Tests of Undercut Anchors,”
published in ACI Special Publication SP-103.
Cannon, R.W., Burdette, E.G., and Funk R.R. (1975), “Anchorage to Concrete," Report
No. CEB 75-32, Tennessee Valley Authority, December 1975.
Cannon, R.W., Godfrey, D.A., and Moreadith, F.L. (1981), “Guide to the Design of
Anchor Bolts and Other Steel Embedments,” Concrete International, Vol. 3, No. 7
(July 1981), pp. 28-41.
Conrad, R.F., (1969), “Tests of Grouted Anchor Bolts in Tension and Shear,” Journal of
American Concrete Institute, Vol. 66, No. 9 (September 1969), pp. 725-728.
Cook, R.A., and Klingner, R.E. (1991), “Behavior of Multiple-Anchor Steel-to-Concrete

Connections with Surface Mounted Baseplates,” Anchors in Concrete - Design
and Behavior SP 130, George A. Senkiw and Harry B. Lancelot III, eds., ACI, pp.
61-122.
DeWolf, J.T., and Ricker, D. (1990), “Design of Column Base Plates,” American
Institute of Steel Construction (AISC), Steel Design Guide Series 1, Chicago, IL.
DeWolf J.T., and Sarisley, E.F. (1980), “Column Base Plates with Axial Loads and
Moments,” Journal of the Structural Division, ASCE 1980; 106(11):2167–84.
94
References
Fahmy, M., Stojadinovic, B., Goel, S.C. (1999), “Analytical and Experimental Studies on
the Seismic Response of Steel Column Bases,” Proceedings of the 8th Canadian
conference on earthquake engineering.
Fenz, D. (2002), “Frictional Properties of Non-Metallic Materials for Use in Sliding

Bearings: An Experimental Study,” Student Research Accomplishments 2001-
2002, pp. 113-118, Department of Civil, Structural & Environmental Engineering,
University at Buffalo, NY.
Fisher, J.M. (1981), “Structural Details in Industrial Buildings,” Engineering Journal,

AISC, Vol. 18, No. 3, Third Quarter, 1981, pp. 83-89.
Fisher, J.M., and Kloiber, L.A. (2006), “Steel Design Guide 1 - Base Plate and Anchor
Rod Design,” 2nd Ed., AISC 801-06, American Institute of Steel Construction,
Inc., Chicago, IL.
Fuchs, W., Eligehausen, R., and Breen, J.E. (1995), “Concrete Capacity Design (CCD)
Approach for Fastening to Concrete,” ACI Structural Journal 92 (1995) (1), pp.
73–94.
Goldman, C. (1983), “Design of Column Base Plates and Anchor Bolts for Uplift and
Shear,” Structural Engineering Practice, Vol. 2, No. 2, pp. 103-115.
Grauvilardell, J.E., Lee, D., Ajar, J.F., and Dexter R.J. (2005), “Synthesis of Design,
Testing and Analysis Research on Steel Column Base Plate Connections in High
Seismic Zones,” Structural engineering report no. ST-04-02. Minneapolis (MN):
Department of Civil Engineering, University of Minnesota.
Gresnigt, A.M., Romeijn, A., Wald, F., and Steenhuis, C.M. (2008), “Column Bases in
Shear and Normal Force,” HERON, Volume 53 (2008), Issue 1/2, Special issue:
Steel Column Bases.
95
References
Kawano, H., Kutani, K., and Masuda, K. (2003), “Experimental Studies on the Shear
Resistance of Anchor Bolts in Exposed Type of Steel Column Bases,” Journal of
Structural and Construction Engineering, Transactions of AIJ, No. 567, May,
2003, pp. 141-148. (In Japanese).
Krawinkler, H., Gupta, A., Medina, R., and Luco, N. (2000), “Loading Histories for
Seismic Performance Testing of SMRF Components and Assemblies,” SAC Joint
Venture, Report no. SAC/BD-00/10. Richmond, CA.
Klingner, R.E., Mendonca, J.A., and Malik, J.B. (1982), “Effect of Reinforcing Details
on the Shear Resistance of Anchor Bolts Under Reversed Cyclic Loading,”
Journal of American Concrete Institute, Vol. 79, No. 1 (January-February 1982),
pp. 3-12.
Kulak, G.L., Fisher, J.W., and Struik, J.H.A. (1987), “Guide to Design Criteria for Bolted
and Riveted Joints,” 2nd edition, John Wiley and Sons, New York, NY.
MacGregor, J., and Wight, J.K. (2004), “Reinforced Concrete: Mechanics and Design,”
4th Edition, Prentice-Hall, Upper Saddle River, NJ.
Nagae, T., Ikenaga, M., Nakashima, M., and Suita, K. (2006), “Shear Friction between
Base Plate and Base Mortar in Exposed Steel Column Base,” Journal of Structural
and Construction Engineering, Architectural Institute of Japan, No. 606, August
2006, pp. 217-223. (In Japanese).
Nakashima, S. (1998), “Mechanical Characteristics of Exposed Portions of Anchor Bolts

in Steel Column Bases under Combined Tension and Shear,” Journal of
Constructional Steel Research, 1998, Vol. 46, No.1-3, Paper No.277. (In
Japanese).
96
References
Occupational Safety and Health Administration (OSHA). (2001), “Safety Standards for
Steel Erection,” (Subpart R of 29 CFR Part 1926), Washington, D.C.
Ožbolt, J., Eligehausen, R., Periškić, G., and Mayer, U. (2007), “3D FE Analysis of
Anchor Bolts with Large Embedment Depths,” Engineering Fracture Mechanics,
74, 168-178.
PCI Design Handbook, First Edition, (1971), “Connections,” Prestressed Concrete

Institute, pp. 6-19, Chicago, IL.
Rabbat, B.G., and Russell, H.G. (1985), “Friction Coefficient of Steel on Concrete or
Grout,” Journal of Structural Engineering Volume 111, No. 3, March 1985, pp.
505–515.
Rotz, J.V., and Reifschneider, M. (1989), “Combined Axial and Shear Load Capacity of
Embedments in Concrete,” 10th International Conference, Structural Mechanics in
Reactor Technology, Aug. 1989, Anaheim, CA.
Scacco, M.N. (1992), “Design Aid-Anchor Bolt Interaction of Shear and Tension Loads,”
Engineering Journal, AISC, Fourth Quarter, 1992/Volume 29, No. 4 pp. 137-140.
Shipp, J.G., and Haninger, E.R. (1983), “Design of Headed Anchor Bolts,”
Engineering Journal, AISC, Vol. 20, No. 2, Second Quarter, 1983, pp. 58-69.
Tronzo, T. M. (1984), “Design of Heavy Steel Column Bases: Handling Forces - Some
Practical Procedures,” Structural Engineering Practice, Vol. 2, No. 4, pp. 279-300.
97
Appendix A: Ancillary Test Data
Appendix A
Ancillary Test Data
Table A.1 - Anchor rod tension test results

Nominal Measured Measured
Young’s Yield Ultimate
diameter original necked
Specimen modulus strength1 strength
(inches) diameter diameter
(ksi) (ksi) (ksi)
(inches) (inches)
1a 0.7483 0.501 32,270 66.81 96.54
3/4
1b 0.7478 0.500 31,700 66.74 96.32
2a 1.2500 0.773 31,220 54.74 74.93
1-1/4
2b 1.2458 0.765 31,050 54.10 74.99
1
Measured yield stress, based on 0.2% offset method
80
Stress, ksi
40
3/4 Inch Diameter Rod - Test 1
3/4 Inch Diameter Rod - Test 2
1-1/4 Inch Diameter Rod - Test 1
1-1/4 Inch Diameter Rod - Test 2
0
0.00 0.01 0.02
Strain, in/in
Figure A.1 – Stress-strain response of all anchor rod tension tests
93
Table A.2 – Concrete cylinder compressive strength test results

Corresponding Cylinder Maximum Twenty-eight
Truck Weight
Specimen large scale test area1 compressive day compressive
number (lbs)
number (in2) load (kips) strength (psi)
1a 28.50 28.274 136.75 4,837
1b 1, 2, 3 & 4 28.75 28.175 120.00 4,259
1c 28.50 28.128 130.00 4,622
2a 28.75 28.161 135.75 4,820
2b 1 5 28.70 28.185 134.25 4,763
2c 28.50 28.119 132.25 4,703
3a 28.50 28.086 138.00 4,913
3b 6 28.50 28.208 134.00 4,750
3c 28.25 28.138 121.00 4,300
4a 28.50 28.152 139.00 4,937
4b 2 7 28.50 28.166 142.00 5,042
4c 28.25 28.138 144.00 5,118
1
Nominal cylinder size = 6” diameter, 12” height
(a) (b)
Figure A.2 – Photographs showing representative concrete

cylinder test (a) before and (b) after failure
94
Table A.3 – Grout cylinder compressive strength test results

Approx. Water
Corresponding Cylinder Maximum Compressive
cure content Weight
Specimen large scale test area1 compressive strength
time (gallons per (pounds)
number (in2) load (kips) (psi)
(days) 50 lbs grout)
A1 4.5 0.875 3.16 7.13 33.25 4,662
A2 4.5 0.875 3.21 7.11 42.00 5,906
4
A3 4.5 0.875 3.26 7.09 50.00 7,057
A4 4.5 0.875 3.23 7.10 49.00 6,904
A5 7.0 0.875 3.21 7.11 50.50 7,104
A6 7.0 0.875 3.18 7.10 42.00 5,916
5
A7 7.0 0.875 3.25 7.10 54.50 7,679
A8 7.0 0.875 3.24 7.12 58.00 8,151
B1 7.0 1.000 3.15 7.07 43.25 6,117
B2 7.0 1.000 3.13 7.10 48.00 6,761
3
B3 7.0 1.000 3.15 7.10 44.00 6,198
B4 7.0 1.000 3.14 7.10 41.50 5,842
B5 9.0 1.000 3.15 7.10 49.00 6,900
B6 9.0 1.000 3.14 7.06 45.50 6,443
1
B7 9.0 1.000 3.16 7.11 35.50 4,996
B8 9.0 1.000 3.12 7.10 34.00 4,788
B9 13.0 1.000 3.09 7.14 51.50 7,216
B10 13.0 1.000 3.18 7.07 56.00 7,917
2
B11 13.0 1.000 3.10 7.11 50.50 7,102
B12 13.0 1.000 3.17 7.09 43.00 6,065
C1 7.0 1.125 3.13 7.08 33.50 4,731
C2 7.0 1.125 3.13 7.10 41.00 5,777
7
C3 7.0 1.125 3.12 7.09 45.00 6,349
C4 7.0 1.125 3.13 7.11 45.00 6,330
C5 9.0 1.125 3.10 7.10 41.00 5,771
C6 9.0 1.125 3.11 7.09 48.00 6,775
6
C7 9.0 1.125 3.12 7.09 35.00 4,933
C8 9.0 1.125 3.12 7.09 47.00 6,625
1
Nominal cylinder size = 3” diameter, 6” height
(a) (b)
Figure A.3 – Photographs showing representative grout

cylinder test (a) before and (b) after failure
95
Figure A.4 – Grout product data from manufacturer (first page)
96
Figure A.5 – Grout product data from manufacturer (second page)
97
Appendix B: Base Plate Test Data
Appendix B
Base Plate Test Data
GUIDE TO BASE PLATE TEST DATA

Frictional force – absolute value of the load recorded by lateral (horizontal) actuators for the
surface friction tests
Lateral displacement – lateral displacement of the base plate measured from the displacement
of the lateral (horizontal) actuators minus lateral displacement of concrete pedestal
Axial force – axial load recorded by vertical (axial) actuators minus the self weight of test
assembly
“Time” – a representation of the progression of the test; since lateral displacement rate is
approximately constant for most tests, this may be considered as the approximate
cumulative lateral displacement
Effective coefficient of friction - frictional force normalized by the axial compression load
Lateral load – load recorded by lateral (horizontal) actuators
Vertical uplift – vertical (axial) displacement of the base plate averaged from four locations
Lateral slip – lateral displacement of the base plate for the shear key tests
Axial Force
Vertical Uplift
Lateral Displacement or Slip

Frictional Force or Lateral Load
Figure B.1 – Schematic illustrating definitions of base plate test data
98
Test #1 Test #1 - 43 kips axial compression

300 40

0 0
-300 -40
-1.1 0.0 1.1 -1.1 0.0 1.1
Lateral Displacement, inches Lateral Displacement, inches
Test #1 - 112 kips axial compression Test #1 - 261 kips axial compression
100 300

0 0
-100 -300
-1.1 0.0 1.1 -1.1 0.0 1.1
Test #1 Test #1
1.1 0
Axial Force, kips
0.0 -150
-1.1 -300
"Time" "Time"
Test #1 Test #1
300 1.0
150 0.5
0 0.0
"Time" "Time"
Figure B.2 – Data for Test #1 – Surface friction with shim stacks
99

300 300

0 0
-300 -300
-1.1 0.0 1.1 -1.1 0.0 1.1

100 40

0 0
-100 -40
-1.1 0.0 1.1 -1.1 0.0 1.1
Test #2 Test #2
1.1 0
Axial Force, kips
0.0 -150
-1.1 -300
"Time" "Time"
Test #2 Test #2
300 1.0
200
0.5
100
0 0.0
"Time" "Time"
Figure B.3 – Date for Test #2 – Surface friction with shim stacks
100

Frictional Force, kips 300 300

0 0
-300 -300
-1.1 0.0 1.1 -1.1 0.0 1.1

100 40

0 0
-100 -40
-1.1 0.0 1.1 -1.1 0.0 1.1
Test #3 Test #3
1.1 0
Axial Force, kips
0.0 -150
-1.1 -300
"Time" "Time"
Test #3 Test #3
300 1.0
200
0.5
100
0 0.0
"Time" "Time"
Figure B.4 – Data for Test #3 – Surface friction without shim stacks
101
40 1

Lateral Load, kips
0
0
-40
-1 0 1 -1
Lateral Displacement, inches "Time"
50 0.6

Axial Load, kips
0 0.3
-50 0.0
"Time" "Time"
Figure B.5 – Data for Test #4 – 3/4” diameter anchor rod bearing
140 1
Lateral Load, kips
0
0
-140
-1 0 1 -1
Lateral Displacement, inches "Time"
125 0.6
100
Axial Load, kips
75
50 0.3
25
-25 0.0
"Time" "Time"
Figure B.6 – Data for Test #5 – 1-1/4” diameter anchor rod bearing
102
200 10
Lateral Load, kips 0
Axial Load, kips

-10
100
-20
-30
0 -40
0.0 0.2 0.4 0.0 0.2 0.4
Lateral Slip, inches Lateral Slip, inches
Figure B.7 – Data for Test #6 – shear key bearing – 5.5” embedment – positive loading direction
200 10
0
Lateral Load, kips
Axial Load, kips

-10
100
-20
-30
0 -40
0.0 0.2 0.4 0.0 0.2 0.4
Figure B.8 – Data for Test #6 – shear key bearing – 5.5” embedment – negative loading direction
200 10
0
Lateral Load, kips
Axial Load, kips
-10
100
-20
-30
0 -40
0.0 0.2 0.4 0.0 0.2 0.4
Figure B.9 – Data for Test #7 – shear key bearing – 3.0” embedment – positive loading direction
200 10
0
Axial Load, kips
Axial Load, kips
-10
100
-20
-30
0 -40
0.0 0.2 0.4 0.0 0.2 0.4
Figure B.10 – Data for Test #7 – shear key bearing – 3.0” embedment – negative loading direction
103
(a)
(b)
(c)
Negative Load
Direction
-------------------
Positive Load
Direction
Figure B.11 – Photograph showing post test condition (negative load direction side extracted) of
pedestal from Test #6 – 5.5” shear key embedment depth (a) elevation view of positive load
direction side (b) elevation view of negative load direction side (c) plan view
104
(a)
(b)
(c)
Negative Load
Direction
-------------------
Positive Load
Direction
Figure B.12 – Photograph showing post test condition of pedestal from Test #7 – 3.0”
shear key embedment depth (a) elevation view of positive load direction side (b)
elevation view of negative load direction side (c) plan view
105
(a)
(b)
(c)
Negative Load
Direction
-------------------
Positive Load
Direction
Figure B.13 – Photograph showing post test condition (failure surfaces extracted) of
pedestal from Test #7 – 3.0” shear key embedment depth (a) elevation view of positive
load direction side (b) elevation view of negative load direction side (c) plan view
106
Figure B.14 – Contour map of Test #6 (5.5” shear key embedment depth - negative load
direction) failure surface recorded by HD laser scanner
107
Appendix C: Anchor Rod Capacity Theory and Examples
Appendix C
Anchor Rod Capacity Theory and Examples
C.1 DEVELOPMENT OF ANCHOR ROD CAPACITY EQUATION CONSIDERING SHEAR AND

TENSION LOADING
According to the American Institute of Steel Construction (AISC, 2005), tests have
shown that the strength of bearing fasteners subject to combined shear and tension
resulting from externally applied forces can be closely defined by an ellipse (Kulak et al.,
1987). The relationship, based on AISC (2005) Equation [C-J3-5a], is expressed as
2 2
⎛ ft ⎞ ⎛ fv ⎞
⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ = 1 [C.1]
⎝ Fnt ⎠ ⎝ Fnv ⎠
Where: Fnt , Fnv = nominal tensile stress and shear stress, respectively
f t , f v = required tensile and shear stress, respectively
From AISC (2005) Equation [C-J3-2], the nominal tensile stress of a bolt is expressed as
Fnt = 0.75 Fu [C.2]
Where: Fu = ultimate tensile stress of bolt
0.75 = a factor accounting for the approximate ratio of the effective area of
the threaded portion of the rod to the area of the shank of the bolt
for common rod sizes
108
From AISC (2005) Equation [C-J3-3], the nominal shear stress, when the threads are
included in the shear plane, is expressed as
Fnv = 0.40 Fu [C.3]
Where: 0.40 = a factor accounting for the effect of shear and for the reduced area
of the threaded portion of the fastener when the threads are not excluded
from the shear plane; AISC (2005) makes this value conservative for the
threaded portion by 11.1% (e.g. the typical shear factor considering the
reduced area is 0.75 * 0.6 = 0.45 ).
The required shear stress is expressed as
V
fv = [C.4]
nA
Where: V = applied shear load

n = number of rods
A = unthreaded area of rod = 0.25 π d 2
d = unthreaded diameter of anchor rod
The required tensile stress, based solely on tensile forces in the rod, is expressed as
P
ft = [C.5]
nA
Where: P = applied axial load
109
Solving Equations [C.1] - [C.5] for shear load provides
(
V = 0.4 0.25 π d 2 n Fu )
2 ⎛ 16 ⎞
− ⎜ ⎟P 2 [C.6]
⎝9⎠
C.2 DEVELOPMENT OF ANCHOR ROD CAPACITY EQUATION CONSIDERING SHEAR AND

TENSION PLUS FLEXURAL LOADING
The strength of bearing fasteners subject to combined shear, tension and flexure loading
is identical to Equations [C.1] – [C.4] except that the required tensile stress is based on
tensile loading plus bending. Thus, the required tensile stress is expressed as
f t = f ta + f tb [C.7]
Where: f ta = the required axial stress from axial loading
f tb = the required axial stress from bending
The required axial stress from the axial load is expressed as
P
f ta = [C.8]
nA
The required axial stress from bending is expressed as
Ml
f tb = [C.9]
nZ
Where: M l = resisting moment of the rod
d3
Z = plastic section modulus of circular cross section =
6
110
The resisting moment of a rod, taken from beam theory, is expressed as
M l = klV [C.10]
Where: k = “effective length” factor of the rod (i.e. lever arm factor)
l = length of rod in bending
Solving Equations [C.1] - [C.5] and [C.7] - [C.10] for shear load provides
V =
(
d 2 12π 2 k l Fu n d ) + (15π F n d ) − (80 P )
2
u
2 2 2
− 64π P k l d
[C.11]
96(π k l ) + 150d
2 2
C.3 VARIATIONS ON THE ANCHOR ROD CAPACITY EQUATION

Two possible variations exist to estimate the capacity of an anchor rod. The first variation
uses the measured, rather than the nominal, anchor rod diameter. Recall in Equation [C.2]
the AISC Steel Construction Manual (2005) uses 75% of the nominal rod area for the
tensile stress area as well as another similar reduction factor for the shear stress area (see
Equation [C.3]). Table 7-18 in the AISC Manual (2005) lists the minimum root diameter
for common bolts. If the minimum root diameter (measured rod diameter) is used, the rod
capacity equation considering tension, shear and flexural stress can be expressed as –
2 2
⎛ 4 P 6klV ⎞ ⎛ ⎞
⎟⎟ = (nFu )2
V
⎜⎜ + 3 ⎟⎟ + ⎜⎜ [C.12]
⎝ π du ⎠ ⎝ 0.15π d u
2 2
du ⎠
Where: du = threaded diameter of anchor rod
111
Additionally, the AISC Steel Construction Manual (2005) states the elliptical relationship
of Equation [C.1] can be replaced by a trilinear relationship as follows –
⎛ ft ⎞ ⎛ fv ⎞
⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ = 1.3
⎝ nt ⎠ ⎝ nv ⎠
F F [C.13]
f t ≤ Fnt ; f v ≤ Fnv
The AISC Manual (2005) describes that this representation offers the advantage that no
modification of either type of stress is required in the presence of fairly large magnitudes
of the other type.
C.4 ANCHOR ROD CAPACITY EXAMPLES

Twelve methods are considered to estimate the predicted strength of the anchor rods in a
grouted base plate connection (with welded plate washers) under shear and tension
loading. The first four estimates of strength ( RnoPVmin al , eliptical , Rmeasued
PV PV
, eliptical , Rno min al , trilinear ,
PV
Rmeasued ,trilinear ) disregards anchor rod bending, but instead considers only the interaction
between axial force and shear; measured and nominal rod diameters, as well as ellipse
and trilinear stress interaction relationships, are considered. The last eight estimates of
strength reflect the interaction of axial, shear and flexure loading in which the anchor rod
is assumed to deform in double curvature. Thus, the “effective length” factor of the rod
(i.e. lever arm factor) “k” is assumed as 0.5. The determination of the bending strength
also relies on an estimate of the length over which this reverse curvature bending occurs.
The AISC Design Guide 1 (Fisher & Kloiber, 2006) assumes that full end fixity is
obtained within the grout pad and rod bolted end, such that the bending length is
calculated as the distance between the top of the grout pad (i.e. bottom of the base plate)
to the center of the welded plate washer. The estimate of strength corresponding to this
bending length, considering measured and nominal rod diameters, as well as ellipse and
trilinear stress interaction relationships, is designated as RnoPVM PVM
min al , eliptical , Rmeasured , eliptical ,
RnoPVM PVM
min al , trilinear or Rmeasured ,trilinear . Based on experimental observations, including evidence
of extensive damage to the grout from cyclic base plate slip which may preclude full end
112
fixity provided by the grout, an alternate bending length is estimated as the distance
between the top of the concrete surface to the center of the welded plate washer. The
estimate of strength corresponding to this bending length (as well as considering
measured and nominal rod diameters, and ellipse and trilinear stress interaction
relationships), is designated as RnoPVM * PVM * PVM * PVM *
min al , eliptical , Rmeasured , eliptical , Rno min al , trilinear , Rmeasured ,trilinear .
Table C.1 tabulates values of the geometry of the base plate connection which details the
rod length in bending for all strength estimates and large scale tests. Furthermore, Table
C.2 lists variables and parameters from the two large scale tests used to calculate the
twelve estimates of anchor rod strength, as well as results of these estimates.
113
Table C.1 – Details of rod bending length

Test Number 4 5
Average thickness of grout (inches) 1.25 1.00
Thickness of base plate (inches) 2.00 2.00
Thickness of the welded plate washer (inches) 0.25 0.5
Length of bending “l” for R PVM estimate1 (inches) 2.125 2.25
Length of bending “l” for R PVM * estimate2 (inches) 3.375 3.25
1
Calculated as the thickness of the base plate plus half the thickness of the welded plate washer
2
Calculated as the thickness of the grout pad plus the thickness of the base plate plus half the thickness of
the welded plate washer
Table C.2 – Details of anchor rod strength estimate calculations

Test Number 4 5
Anchor rod unthreaded diameter “d” (inches) 0.75 1.25
Anchor rod threaded diameter “du” (inches) 0.64 1.10
Ultimate tensile stress of rod “Fu” 1 (ksi) 96.43 74.96
Number of rods “n” 4 4
Effective length factor “k” 0.5 0.5
Imposed axial load “P” (kips) 39.6 108
Rod strength “ RnoPVmin al , eliptical ” as per Equation [C.6] (kips) 64.8 135
2
Rod strength “ RnoPVM
min al , eliptical ” as per Equation [C.11] (kips) 12.9 37.5
3
Rod strength “ RnoPVM *
min al , eliptical ” as per Equation [C.11] (kips) 8.23 26.7
PV
Rod strength “ R measued , eliptical ” as per Equation [C.12] (kips) 71.1 158
PVM 2
Rod strength “ R measured , eliptical ” as per Equation [C.12] (kips) 10.8 35.5
PVM * 3
Rod strength “ Rmeasured , eliptical ” as per Equation [C.12] (kips) 6.84 25.0
Rod strength “ RnoPVmin al ,trilinear ” as per Equation [C.13] (kips) 65.7 134
2
Rod strength “ RnoPVM
min al , trilinear ” as per Equation [C.13] (kips) 13.2 39.6
PVM * 3
Rod strength “ R no min al , trilinear ” as per Equation [C.13] (kips) 8.31 27.4
PV
Rod strength “ R measued ,trilinear ” as per Equation [C.13] (kips) 73.6 157
PVM 2
Rod strength “ Rmeasured ,trilinear ” as per Equation [C.13] (kips) 10.9 6.88
PVM * 3
Rod strength “ Rmeasured ,trilinear ” as per Equation [C.13] (kips) 36.7 25.4
Observed anchor rod peak strength, positive direction “ R peak ” (kips) 30.2 126
reversed
Observed anchor rod peak strength, reverse direction “ R peak ” (kips) 28.2 70.4
1
Determined by ancillary tension tests of the anchor rods
2
Using a length of bending equal to 2.125” for Test #4 and 2.25” for Test #5
3
Using a length of bending equal to 3.375” for Test #4 and 3.25” for Test #5
114
Appendix D: Analysis of Anchor Rod Response
Appendix D
Analysis of Anchor Rod Response
D.1 INTRODUCTION
Recall from Chapter 3 that the load-deformation response of the base plate tests
investigating anchor rod strength (Tests #4 and #5) is highly complex. In addition, the
response does not demonstrate a well defined peak. Thus, it is somewhat challenging to
characterize the strength capacity associated with anchor rod failure. The strength is
influenced by the interaction of several phenomena, including (1) increased tension
forces of the rods due to large deformations (i.e. geometric second order effects) (2)
strength and stiffness degradation due to grout damage and (3) anchor rod contact within
the base plate holes (rather than with the welded plate washers). In addition to the
difficulty in measuring a peak capacity, these complex interactions necessitate a
subjective (rather than a quantitative) evaluation of strength prediction approaches, since
distinct failure modes are not observed. Motivated by these issues, this Appendix presents
a simplified nonlinear analysis of the anchor rods, incorporating the effects of axial force,
moment and shear interaction as well as geometric nonlinearities introduced by
deformation of the anchor rods under large displacements. The main objectives of this
analysis are to –
1. Complement experimental data and visual observations by developing an

understanding of the key factors that affect the response of anchor rods under the
combined actions of axial stress, bending and shear.
2. Develop insights into the relative contributions of the various phenomena that
influence response.
3. Examine the influence of important parameters, such as the effective bending

length on the response.
115
D.2 SIMPLIFIED NONLINEAR ANALYSIS OF ANCHOR ROD STRENGTH UNDER COMBINED
AXIAL FORCE, BENDING AND SHEAR
Figure D.1 schematically illustrates a deformed anchor rod within a base plate under
shear loading. The rod is assumed to experience double curvature bending as observed
experimentally (recall prior discussion in Chapter 3). Figure D.2 presents a detailed
schematic of the anchor rod free-body diagram with geometric parameters such as the
lateral displacement Δ (relative to the concrete footing), the length of bending Leffective and
an effective angle of rod rotation θ, which can be approximately 1 expressed as –
θ = arctan ( Δ / L ) [D.1]
Figure D.4 shows the free-body diagram of half of the bending length; V1 and P1 are the
resultant shear and axial forces, respectively, at the inflection point of the rod. These
resultant cross sectional shear and axial forces may be determined from the applied
lateral load V and tensile load P by the following equations –
V1 = V cosθ − P sin θ
[D.2]
P1 = V sin θ + P cos θ
Thus, the applied shear stress at the cross section may be expressed as –
V1
fv = [D.3]
nA
At = threaded area of rod = 0.25 π dt

2
Where:
dt = threaded diameter of anchor rod
1
Assuming that a plastic hinge forms at the ends of the rods, whereby the rod is approximately aligned
with its chord
116
The applied axial stress due to the axial force may expressed as –
P1
f ta = [D.4]
nAt
The axial stress from bending may be expressed as –
Ml
f tb = [D.5]
nZ t
Where: M l = moment resisted by the rod

3
dt
Z t = plastic section modulus of threaded cross section =
6
Based on equilibrium on the assumed deformed anchor rod shape, the resisting moment
of a rod may be expressed as -
(
M l = V × Leffective − P × Δ / 2 ) [D.6]
An adaptation of the AISC (2005) interaction equation for bearing fasteners (also
2
referenced in the AISC Steel Design Guide Series 1 [Fisher & Kloiber, 2006]) is
expressed as –
2 2
⎛ f ta + f tb ⎞ ⎛ fv ⎞
⎜⎜ ⎟⎟ + ⎜⎜ ⎟⎟ = 1 [D.7]
⎝ Fnt ⎠ ⎝ 0.6 Fnt ⎠
Substituting Equations [D.1]-[D.6] into Equation [D.7] provides an equation which may
be solved to generate a relationship between the applied shear force and the relative
2
A trilinear relationship is often used as an approximation for the elliptical interaction equation presented
here. The elliptical equation is utilized in this section.
117
deformation between the two ends of the anchor rod, i.e. V = f (Δ) . While a closed form
solution is available, it is very large and therefore is not presented. The resulting
relationship represents a plastic limit state (an upper bound solution) describing the load
versus displacement response of the anchor rods. Figures D.4 and D.5 illustrate these
relationships overlaid on the experimental response for Tests #4 and #5, respectively.
Referring to the figures, plastic limit state solutions are generated for three effective
bending lengths, i.e. (1) the distance from the surface of the concrete to the middle of the
plate washer, i.e. Leffective = t plate + twasher / 2 (2) the distance from the top of the grout pad to
the middle of the plate washer, i.e. Leffective = t plate + twasher / 2 + t grout and (3) the distance from
the surface of the concrete to the bottom of the base plate, i.e. Leffective = t grout . The
intersection of these solutions with the vertical axis (i.e. at Δ = 0 ) represent the strength
capacities calculated without the consideration of second order effects, such as described
in Chapter 3 and Appendix C. Also overlaid on the plots is the initial elastic stiffness,
which is expressed as –
12 EI t
kinitial = [D.8]
Linitial 3
Where: E = Young’s modulus
π dt 4
I t = second moment of area of threaded cross section =
64
Linitial = Effective bending length of the undeformed anchor rod, equal to
the thickness of the base plate plus half the thickness of the plate
washer
D.3 OBSERVATIONS FROM NONLINEAR ANALYSIS

A comparison of experimental data with respect to the plastic limit solutions generated in
the previous section provides several useful insights into anchor rod response. Specific
observations for each experiment are first outlined, before summarizing general
118
observations. Referring to Figure D.4, several observations may be made regarding the
response of Test #4 –
1. The initial strength achieved by Test #4 is higher than predicted by any of the
three plastic limit lines. Discussed previously in Chapter 3, all four holes in Test 4
contained some amount of grout. As expected, the presence of this grout restrains
the anchor rods, resulting in very short bending lengths and correspondingly
higher strengths.
2. Once this grout had broken and had been crushed by repetitive cyclic loading, the
strength and stiffness of the experimental response degrades, possibly indicating
that the rod loosens the grout around it and bends/yields over a progressively
increasing length. It is interesting to note that the experimental hysteresis
envelopes in the displacement range (-0.4 inch to +0.4 inch) are approximately
coincident with the limit lines corresponding to Leffective = t plate + twasher / 2 for the
initial cycles and Leffective = t plate + twasher / 2 + t grout for the subsequent cycles. This
indicates that the strength degradation may be attributed to an increase in effective

bending length of the anchor rods.
3. At larger displacements in both directions of loading (-0.5 inches and +0.5

inches), a significant increase in stiffness is observed. Several interesting
observations may be made regarding this increased stiffness –
a. This response is not captured by the second-order analysis, which

considers tension stiffening in the deformed anchor rod. In fact, the second
order analysis solution predicts similar abrupt stiffening at displacements
of approximately 2 inches (outside the limits of Fig. D.4). The second
order analysis is dominated by “ tan θ ” term, which appears linear for
θ < 400 . For instance, the peak deformations observed in the tests are on
the order of 15-20 degrees.
119
b. The increase in stiffness, observed in the experiment, occurs at different

displacements in the positive and negative directions. On the other hand,
tension stiffening predicted by the analysis does not predict this type of
unsymmetrical response.
c. The increased stiffness is roughly equal to that predicted by the limit

solution corresponding to Leffective = t grout
The three points discussed above indicate that the increase in stiffness may be
attributed to the base plate impinging (i.e. making contacting) upon the anchor
rod, thereby reducing the bending length, rather than tension stiffening (which
would become dominant at significantly larger deformations and exhibit
symmetric response with respect to the displacements).
Recall that grout was completely removed from the base plate holes in Test #5. However,
in Test #5, the clearance between the anchor rods and hole was smaller as compared to
Test #4 (due to the larger rod diameter). Moreover, rods were placed with a larger degree
of eccentricity with respect to the center of the hole, such that they were very close
(almost flush) to the edge of the base plate hole in one loading direction (corresponding
to the positive quadrant of Fig. D.5). Referring to Fig. D.5 and prior discussion in
Chapter 3, several observations may be made regarding the response of Test #5 –
1. In the “positive” loading direction (with respect to Fig. D.5), the limit solution
corresponding to Leffective = t grout results in a reasonably accurate prediction of the
envelope response. This, along with measurements of the rod placement, suggests
that once the base plate made contact with some (or all) rods, the rod was
subjected to flexure below the surface of the base plate, i.e. within the thickness
of the grout.
2. Conversely, in the opposite loading direction (where a larger clearance was

available between the plate and the anchor rods) the limit solution corresponding
120
to Leffective = t plate + twasher / 2 provides a good initial estimate of the observed
response.
3. Progressive cyclic loading degraded the strength as well as the stiffness of the
specimen, such that the degraded envelope (prior to the plate making contact with
the rod) in both directions is approximated by the limit line corresponding to
Leffective = t plate + twasher / 2 + t grout .
4. Similar to Test #4, stiffening behavior (at larger displacements) is observed in

Test #5. However, for Test #5, the response is highly unsymmetrical – consistent
with the unsymmetrical placement of the anchor rods within the holes.
5. The “negative” loading direction of Test #5 provides the best example of anchor
rod bending not affected by factors such as grout (in the anchor rod holes)
constraining the rod or contact with the base plate. Thus, this response is
characterized by unrestrained double curvature bending contacting the welded
plate washer. The response in this loading direction is represented by the limit
solution corresponding to Leffective = t plate + twasher / 2 for the initial loading cycles,
and subsequently is closely followed by the limit solution corresponding to

Leffective = t plate + twasher / 2 + t grout , indicating the gradual increase of effective length,
possibly due to grout damage or rod elongation.
Based on the above discussion, the following general observations are presented, with a
view to provide speculative insights into probable progression of events during the
experiments –
1. At the start of testing, the strength of the specimen is controlled by bending of the
anchor rod over the effective length Leffective = t plate + twasher / 2 , disregarding effects
such as early contact between the base plate and anchor rod or presence of grout
in the holes.
121
2. Under subsequent cyclic deformations, the anchor rod bending length increases
until it loosens/damages all the grout around it. At this time, the envelope
response may be described based on Leffective = t plate + twasher / 2 + t grout .
3. In either loading direction, if the base plate (rather than the plate washer) makes
contact with the anchor rod, the bending length is abruptly reduced (resulting in
Leffective = t grout ), accompanied by a rapid increase in force.
4. The abrupt increase in force (although consistent with tension stiffening) is

attributed to contact between the base plate and anchor rod. This may be
explained based on two factors. (1) Tension stiffening response will be
symmetric, whereas the observed response is unsymmetrical. The increase in
force due to the base plate impinging on the anchor rod depends on the placement
of the anchor rod, and the corresponding response will, in general, be asymmetric.
(2) The nonlinear analysis model presented in this Appendix predicts an abrupt
increase in force due to tension stiffening at significantly greater displacements as
compared to those observed in the experiments.
122
Undeformed Rod
anchor rod
nut
welded
washer
base plate
grout
concrete
Plate Washer Bearing Damage to Grout

L = half plate L = half plate
washer thickness washer thickness
plus base plate bearing location plus base plate and
thickness grout thickness
load load
damage to grout
Tension Stiffening Plate Bearing
load load
tensile force in rod L = grout thickness bearing location
Figure D.1 – Schematic illustrating various phenomenon in the anchor rod bearing
mechanism
123
Plastic hinge
Inflection
point
Leffective
Deformed
anchor rod
90º - θ
Plastic hinge
Δ
V
M
P
Figure D.2 – Free-body diagram of the deformed anchor rod
124
P P1
Deformed
anchor rod
V
θ
V1
L/2
90º - θ
Δ/2
V
M
P
Figure D.3 – Free-body diagram of half of the deformed rod
125
50
Test Elastic response – Leffective = tplate+ ½ twasher
Elastic response (plate + 1/2 washer)
plate + 1/2 washer
plate+grout+1/2washer
grout only
25
PVM – Leffective = tgrout
PVM – Leffective = tplate+ ½ twasher

Force (kips)
PVM – Leffective = tplate+ ½twasher + tgrout
0
-1 -0.5 0 0.5
PVM – Leffective = tplate + ½twasher + tgrout 1
-25 PVM – Leffective = tgrout
-50
Displacement (inches)
Figure D.4 – Experimental and predicted response of Test #4 (3/4” diameter anchor rods)
150
Test Elastic response – Leffective = tplate+ ½ twasher
Elastic response (plate + 1/2 washer)
plate + 1/2 washer
plate+grout+1/2washer
grout only
100
50
Force (kips)
0
-1 -0.5 0 0.5 1
-50 PVM – Leffective = tplate+ ½ twasher
-100
Displacement (inches)
Figure D.5 – Experimental and predicted response of Test #5 (1-1/4” diameter anchor rods)
126
Appendix E: Shear Key Capacity Theory and Examples
Appendix E
Shear Key Capacity Theory and Examples
E.1 DEVELOPMENT OF THE CCD SHEAR KEY CAPACITY EQUATION

Based on the concrete capacity design (CCD) approach for fasteners in concrete
(Equation [9a] in Fuchs et al., 1995), the concrete cone failure load of a single anchor
loaded in tension in uncracked, unreinforced concrete unaffected by edge influences or
overlapping cones of neighboring anchors is given by
N no = k nc f c′ hef1.5 , lb [E.1]
Where: k nc = calibration factor = 40 for cast-in situ headed studs and headed
anchor bolts
f c′ = concrete compression strength measured on 6 by 12 inch cylinders,
psi
hef = effective embedment depth, inches
Assuming an idealized (i.e. pyramid shaped) concrete cone with an inclination between
the failure surface and the surface of the concrete member of about 35 degrees, the
concrete cone failure load can be related by
⎛ 1 ⎞
N no = 40 f c′ hef1.5 = k 35 ⎜ ⎟A f c′ [E.2]
⎜ h ⎟ no
⎝ ef ⎠
Where: k 35 = constant
Ano = projected area of one anchor at the concrete surface unaffected by
edge influences or neighboring anchors, idealizing the failure cone

as a pyramid with a base length of 3 hef = 9 hef2
127
Solving Equation [E.2] for k 35 provides a constant of 40 / 9 = 4. 4 . Thus, the concrete

cone failure load can be expressed as
40 ⎛⎜ 1 ⎞⎟
N no = Ano f c′ , lb [E.3]
9 ⎜ hef ⎟
⎝ ⎠
Shear lugs typically have a very large lateral stiffness such that the shear capacity of a
shear lug in concrete loaded towards an edge may be interpreted in a similar fashion as an
anchor loaded in tension. Thus, the shear load capacity of a shear key loaded towards a
free edge may be expressed as
40 ⎛ 1 ⎞
VnCCD = ⎜ ⎟ A35 f c′ , lb [E.4]
9 ⎜⎝ c ⎟⎠
Where: c = free edge distance from the shear key, inches

A35 = effective area based on a 35 degree projected plane from the bearing
edges of the shear lug to the free surface, excluding the bearing area
of the shear lug, inches2
E.2 EXAMPLE CALCULATION OF SHEAR KEY CAPACITY USING THE CCD METHOD
The geometry of the concrete pedestal used to represent the foundation for the base plate
tests is illustrated schematically in Figure E.1. The pedestal measures 48 inches by 84
inches in area and 32.5 inches in height. For both tests, the shear key was installed in the
center of the concrete block such that an edge distance of 20.25” was provided from the
shear key bearing surface to the pedestal edge. Two shear key embedment lengths,
measured from below the surface of the concrete, were tested: 5.5” for Test #6 and 3.0”
for Test #7. The shear key itself was 6” wide for both tests, thus providing a 33 in2
bearing surface for Test #6 and 18 in2 for Test #7. Note that for both tests, the projected
effective failure area of the shear key is limited by the pedestal height. Schematics
128
detailing the calculation of the projected effective failure area for the CCD method are
illustrated in Figure E.2.
The effective area A35 for Test #6 is calculated as
A35 = (height ) ⋅ (width ) − (bearing area of the shear lug )

= (32.5") ⋅ ((1.5 + 1.5) ⋅ 20.25"+6") − (6") ⋅ (5.5") = 2,136 inches 2
A35 = (32.5") ⋅ ((1.5 + 1.5) ⋅ 20.25"+6") − (6") ⋅ (3.0") = 2,151 inches 2
Table E.1 lists all pertinent variables and parameters from the two shear lug tests used to
calculate the shear key capacity using the CCD method.
E.3 EXAMPLE CALCULATION OF SHEAR KEY CAPACITY USING THE 45 DEGREE CONE
METHOD
The 45 degree cone method is prescribed by ACI 349 (2006) for the concrete shear
capacity of embedded shear lugs and is featured in the AISC Design Guide 1 (Fisher &
Kloiber, 2006). As per this method, the strength capacity of the concrete corresponding to
shear blowout is determined as follows –
( )
Vn45 = 4 f c′ A45 , lb [E.5]
Where: f c′ = concrete compressive strength, psi
A45 = effective area based on a 45 degree projected plane from the bearing
edges of the shear lug to the free surface, excluding the bearing area
of the shear lug, inches2
129
Thus, the method assumes a uniform tensile stress of 4 f c′ acting on an effective stress
area defined by projecting a 45 degree plane from the bearing edges of the shear lug to
the concrete free surface. The bearing area of the shear lug is excluded from the projected
area. Note that for both tests, the projected effective failure area of the shear key is not
limited by corner influences or pedestal thickness. Schematics detailing the calculation of
the projected effective failure area for the 45 degree cone method are illustrated in Figure
E.3.
A35 = (height ) ⋅ (width ) − (bearing area of the shear lug )

= (20.25"+5.5") ⋅ ((2 ) ⋅ 20.25"+6") − (6") ⋅ (5.5") = 1,164 inches 2
A35 = (20.25"+3.0") ⋅ ((2 ) ⋅ 20.25"+6") − (6") ⋅ (3.0") = 1,063 inches 2
Table E.1 lists all pertinent variables and parameters from the two large scale tests used
to calculate the shear key capacity using the 45 degree cone method.
130
Table E.1 – Details of shear key capacity calculations

Test Number 6 7
Pedestal length (inches) 84 84
Pedestal width (inches) 48 48
Pedestal height “h” (inches) 32.5 32.5
Corner distance “e” (inches) 39 39
Edge distance “c” (inches) 20.25 20.25
Shear key width “w” (inches) 6.0 6.0
Shear key embedment depth “d” (inches) 5.5 3.0
1
Concrete compressive strength “ f c′ ” (ksi) 4,650 5,030
2
Effective area based on a 35 degree projected plane “ A35 ” (inches2) 2,136 2,151
2 2
Effective area based on a 45 degree projected plane “ A45 ” (inches ) 1,164 1,063
Shear capacity “ Vn
CCD
” calculated as per Equation [E.4] (kips) 144 151
45
Shear capacity “ V ” calculated as per Equation [E.5] (kips)
n 318 302
1
Determined by the average of ancillary compressive tests of concrete cylinders
2
Calculated by projecting planes from the bearing edges of the shear lug to the concrete free surface,
excluding the bearing area of the shear lug and considering corner influences and pedestal thickness; see
section E.2 and E.3 for details
131
shear load
c = 20.25”
Test #6: d = 5.5”
Test #7: d = 3.0” h = 32.5”
w = 6”
84”
48”
Figure E.1 – Schematic of concrete pedestal geometry
132
shear load
c
A35 = (d + 1.5c) * (w + 3c) – (d * w)
w
35º = 1.5c * (w + 2d + 3c)
d
d + 1.5c
A35
w + 3c
A35 = 0.7[(d + 1.5c) * (w + 1.5c + e) - (d * w)]

shear load
c if: e ≤ 1.5c
e 35º
w The 0.7 factor considers the disturbance of

d d + 1.5c the symmetric stress distribution caused by
a corner (Fuchs et al., 1995)
A35
0.5w +1.5c
0.5w + e
shear load
c A35 = (h) * (w + 3c) – (d * w)
if: h ≤ 1.5c + d
w
35º h
d
A35
w + 3c
Figure E.2 – Schematic illustrating the effective stress area for the CCD
method under various boundary conditions (c = edge distance; w = shear key
width; d = shear key embedment depth; e = corner distance; h = foundation
height)
133
shear load
c
A35 = (d + c) * (w + 2c) – (d * w)
= c * (w + 2d + 2c)
w
45º
d d+c
A45
w + 2c
A35 = (d + c) * (w + c + e) – (d * w)
shear load
c if: e ≤ c
e 45º
w
d d+c
A45
0.5w + c
0.5w + e
shear load A35 = (h) * (w + 2c) – (d)*(w)

c
if: h ≤ c + d
w h
45º
d
A45
w + 2c
Figure E.3 – Schematic illustrating the effective stress area for the 45 degree
cone method under various boundary conditions (c = edge distance; w = shear
key width; d = shear key embedment depth; e = corner distance; h = foundation
height)
134

Shear Transfer in Exposed Column Base Plates PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Shear Transfer in Exposed Column Base Plates PDF

Uploaded by

Copyright:

Available Formats

2. Background and Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.2 Current Design Provisions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.3 Current State of Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.3.1 Surface Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.3.2 Anchor Rod Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.3.3 Shear Key Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.4 Objectives of Current Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3. Experimental Program and Test Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.2 Ancillary Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3.2.1 Anchor Rod Tension Tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3.2.2 Standard Concrete Test Cylinders. . . . . . . . . . . . . . . . . . . . . . . 16

3.2.3 Grout Test Cylinders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.3 Large Scale Experiments: Test Setup and Preparation . . . . . . . . . . . . . . 18

3.3.1 Test Setup. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.3.1.1 Test Setup for the Friction Tests:

Tests #1, #2 and #3. . . . . . . . . . . . . . . . . . . . . . . . . . 19

Tests #4 and #5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.3.1.3 Test Setup for the Shear Key Tests:

Tests #6 and #7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.3.2 Loading Protocol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.4 Large Scale Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.4.1 Surface Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.4.2 Anchor Rod Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.4.3 Shear Key Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4. Summary and Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

4.1.1 Surface Friction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

4.1.2 Anchor Rod Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

4.1.3 Shear Key Bearing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

4.2 Conclusion and Design Implications. . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

Appendix A – Ancillary Test Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Appendix B – Base Plate Test Data. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Appendix C – Anchor Rod Capacity Theory and Examples. . . . . . . . . . . . . . . . . . . . . 108

C.1 Development of Anchor Rod Capacity Equation Considering

C.3 Variations on the Anchor Rod Capacity Equation. . . . . . . . . . . . . . . . . . 111

C.4 Anchor Rod Capacity Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Appendix D – Analysis of Anchor Rod Response. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

D.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

D.2 Simplified Nonlinear Analysis of Anchor Rod Strength Under

D.3 Observations from Nonlinear Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Appendix E – Shear Key Capacity Theory and Examples . . . . . . . . . . . . . . . . . . . . . . 127

E.1 Development of the CCD Shear Capacity Equation. . . . . . . . . . . . . . . . 127

E.2 Example Calculation of Shear Key Capacity Using the CCD

E.3 Example Calculation of Shear Key Capacity Using the 45

3.1 Summary of anchor rod tension tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.2 Summary of concrete compression tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.3 Summary of grout compression tests. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.4 Base plate test matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.5 Base plate test loading details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

3.6 Base plate surface friction test results . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . 48

3.7 Anchor rod bearing test results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.8 Shear key bearing test results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

A.1 Anchor rod tension test results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

A.2 Concrete cylinder compressive strength test results. . . . . . . . . . . . . . . . . . . . . . . 94

A.3 Grout cylinder compressive strength test results. . . . . . . . . . . . . . . . . . . . . . . . . 95

C.1 Details of rod bending length. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

C.2 Details of anchor rod strength estimate calculations. . . . . . . . . . . . . . . . . . . . . . 114

E.1 Details of shear key capacity calculations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

3.3 Steel shear loading assembly for base plate tests. . . . . . . . . . . . . . . . . . . . . . . . . 52

3.6 Post-test photograph of grout pad for Test #3 . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.8 Schematic plan view of base plate specimen (to scale). . . . . . . . . . . . . . . . . . . . . 55

3.10 Welded shear key (shown here for Test #6). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3.11 Schematic detail of shear key geometry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3.17 Frictional force versus cumulative slip for Test #2. . . . . . . . . . . . . . . . . . . . . . . . 62